Comparative Morphology of Idiomyces and its Possible Allies Diplomyces, Sandersoniomyces, Symplectromyces, and Tetratomyces (Ascomycetes: Laboulbeniales)

Pertinent stages of development of the receptacle, primary and secondary appendages, and perithecium of Idiomyces peyritschii are described and illustrated and comparison is made between this species and the type species of Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces for which early stages of receptacular and appendage development also are described and illustrated. Possible close relationship of these genera with one another is examined in the light not only of morphological similarities and differences but also of host range. It is concluded that Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces do indeed constitute a closely allied group of genera. However, the gap between this alliance and Idiomyces is deemed great enough to warrant their separation into separate tribes within the Laboulbeniales, proposed by Thaxter in 1908, i.e., ldiomyceteae for Idiomyces and Teratomyceteae for the other four genera.


INTRODUCTION
Idiomyces peyritschii was described by Thaxter (1893) as the type of a new genus of Laboulbeniales parasitizing a staphylinid beetle, De/easter dichrous (Grav.),collected in Germany.Thaxter named the species for the pioneer student ofLaboulbeniales, J. Peyritsch, who, perhaps in error, listed D. dichrous as one of the hosts for his newly described Laboulbenia vulgaris Peyr. (1873:249).Peyritsch later (1875:385) claimed also to have observed on D. dichrous a second species of Laboulbenia Mont.& Robin (Robin 1853), i.e., L. gracilis (non L. gracilis Speg.[1915]), which, however, he never described.Laboulbenia vulgaris occurs on various Carabidae, especially species of Bembidion Latreille, several of which were cited as hosts by Peyritsch.No species of Laboulbenia has subsequently been found on De/easter Erichson spp., and Peyritsch may have misinterpreted the true nature of immature specimens of/.peyritschii which vaguely resemble Laboulbenia and as a result may have assigned them to the common L. vulgaris.
Based on the arrangement of antheridia on fertile appendages, Thaxter ( 1893Thaxter ( , 1896) ) suggested that Idiomyces might be most nearly allied to Stigmatomyces Karsten (1869) despite the marked differences between these genera in the structure of their receptacles and in the manner of origin of their appendages and perithecia.In 1908, however, he noted the resemblance of Idiomyces to "the group of genera immediately associated with Teratomyces [Thaxt.],"i.e., Diplomyces Thaxt.( 18 9 5) and Symplectromyces Thaxt.(1908).However, lack of information regarding early stages of development of these genera prevented a critical comparison of I. peyritschii with them for indications of close relationship, at least on morphological grounds.
In 1968, I described Sandersoniomyces divaricatus, type of a new genus, and compared it with Diplomyces actobianus Thaxt.(1895), Symplectromyces vulgaris (Thaxt.)Thaxt. ( 1908; syn.Teratomyces vulgaris Thaxt. [1900]), and T. miri.ficusThaxt.(1893), the type species of their respective genera.Sandersoniomyces is morphologically most like Diplomyces, but all four genera share similarities that suggest their close relationship.My study did not include Idiomyces, and my purpose here is to describe and compare the development of I. peyritschii with that of the above species and, hopefully, determine whether or not Idiomyces should be closely allied with this group of genera.

MATERIALS AND METHODS
For this and my previous study (Benjamin 1968a), observations were made on and illustrations were prepared from among the following collections, all of which along with their hosts originally had been preserved in 50-70% alcohol.All of the fungi are mounted in glycerine containing a trace of cotton blue or acid fuchsin dyes.The slides are in RSA.

OBSERVATIONS A. Development ofldiomyces peyritschii
The ascospore of Jdiomyces peyritschii is two celled like that of all Laboulbeniales.The hyaline sheath surrounding the body of the spore is conspicuously enlarged distally around the longer of the two cells (Fig. 1, 25).The longer cell, originally oriented towards the perithecial ostiole prior to spore discharge (Fig. 1 7), forms the foot (Fig. 2) and eventually gives rise to all cells of the receptacle, numerous secondary appendages, and perithecia (18)(19)(20)(21).The tip of the lower cell is separated as a hyaline, enucleate, homogeneous end-piece that contrasts sharply with the granular contents of the adjacent cell (Fig. 1, 25) and appears somewhat sheathlike in character.
Among the 178 preperithecialjuvenile individuals available for study, the earliest stage in thallus development that I could find is shown in Fig. 2 where the foot is differentiating but divisions of the lower and upper cells, separated by the original spore cross wall (a), have not begun.
In the stage of development shown in Fig. 3, divisions of the upper spore segment have resulted in three superposed cells.This is the young primary appendage (pa) and for purposes of further reference the lower two cells are designated by the letters band c; the upper cell is terminated by a spine (sp) representing the indurated spore apex.This spine typically persists as the primary appendage matures 18).
The lower spore segment (Fig. 3) divides and forms the primary receptacle consisting of three superimposed cells, I, II, and III (Fig. 4).Cell II of the young receptacle soon is divided into two cells, Ila and lib, by an oblique cross wall (Fig. 5).Cell Ila then is separated by a transverse cross wall into two cells, Ila and Ila' (Fig. 6-7), the latter cell more or less paralleling cell Fig. 1-17.ldiomyces peyritschii.-I.Ascospore showing conformation of hyaline adhesive sheath, cross wall (a), and hyaline basal end-piece.-2-10.Pertinent stages of development of the receptacle, primary appendage (pa), and primary and secondary receptacular upgrowths (pru, sru).Details and terminology are given in the text.-11.Young individual prior to appearance of first perithecial initial.Note relationship of primary appendage (pa) and primary and secondary receptacular upgrowths (pru, sru) to cells of the primary receptacle (I, Ila, Ila', lib).Both receptacular upgrowths are forming secondary appendages.-12.Young secondary appendage.-13.Secondary appendage showing basal and terminal sterile branchlets and two tiers of antheridia.-14.Young perithecial initial (see Fig. 18,inset).Basal cell m has been formed by the primary stalk cell, VI; the secondary stalk cell, VII, already has cut off basal cell n and is growing upward around the near face of cell n and will eventually cut off basal cell n'.Also shown are the carpogenic (cp) and trichophoric (!c) cells of the young female organ.-15.Young perithecium showing well-developed trichogyne (tr) (see Fig. 19).The secondary stalk cell, VII, and basal cell mare beginning to elongate.The first two cells (o, w) of the outer rows of wall cells have been formed (those derived from basal cells m and n' are shown) and are growing upward around the female sexual organ and the first-formed inner wall cells.-16.Still later stage of perithecial development as viewed from the side opposite that shown in Fig. 15.Three wall cells (o, w, w') have now been delimited in each row of outer wall cells.Basal cell n has given rise to two of these wall cell rows.Cells VII and m are continuing to elongate.The trichogyne has disappeared except for a small basal remnant.-17.Mature perithecium less most of the primary stalk cell.There are four tiers of outer wall cells (w'-w').Basal cell n' is relatively small compared to the greatly elongate cells VII and m. (Fig. 1

II II
A B CD 1 ~ lib (Fig. 7) which is regarded as posterior in position.As the receptacle continues to grow and mature, cells I, Ila, and lib typically remain undivided (18)(19)(20)(21)(22), although adventitious septa occasionally are formed in one or more of these cells.
Development of the primary appendage is illustrated in Fig. 5-10.As shown in Fig. 5-6, the original spore septum (a) and the septum separating cell b from cell c become blackened.The terminal spinose cell elongates and divides and eventually forms a simple branch consisting of several cells, the lower of which bears the divaricate, indurated spore apex .
A second sterile branch subtending the spinose cell arises from cell c (Fig. 6-11).Finally, development of the primary appendage is completed by the formation of a simple or divaricate, sterile, cellular branchlet from cell b (Fig. 7-11).
Continued development of the receptacle involves the formation of a progression of cells, the primary receptacular upgrowth (pru) from cell III (Fig. 8-9).The earliest stage of formation of these cells is seen in Fig. 8 where a cell cut off by cell III already is forming an upright branch.In the young individual shown in Fig. 9 three nearly vertical cells have been delimited from cell III.The basal cell of the secondary appendage arising from the outer cell, as in most appendages formed by /. peyritschii, is delimited above and below by black septa.
As growth and enlargement of the thallus proceeds, the primary receptacular upgrowth assumes a nearly vertical position concomitant with the development of what can be termed the secondary receptacular upgrowth (sru) from cell Ila' opposite cell lib (Fig. 11,(18)(19)(21)(22)(23).The primary appendage gradually is deflected aside and projects laterally from the base of the primary receptacular upgrowth (Fig. 11,18) where it usually is obscured by antheridia1 branches and commonly is broken or missing in older thalli .
The several cells of the primary receptacular upgrowth cut off smaller cells to the right and left, and each of these cells may form additional cells.As they appear, each cell gives rise to a simple or branched, sometimes sterile but mostly fertile, several-celled secondary appendage (Fig. 11,(18)(19)(20).Cells comprising sterile appendages or sterile branchlets of fertile appendages typically are two to many times longer than broad; fertile cells destined to delimit antheridia are nearly isodiametric or slightly wider than high (Fig. 12-13).
Secondary appendages vary in number of fertile cells and in the number ofbranchlets they may form.The basal cell, as already mentioned, typically is delimited above and below by blackened septa.This cell may occasionally give rise to antheridia or a sterile branchlet (Fig. 13).The axis of a fertile appendage above the basal cell may be 8-10 cells in length.The terminal cell and frequently the preceding one or two cells may develop short or elongate, several-celled prolongations (Fig. 13).Each fertile cell cuts off a single cell from its inner face which forms a simple, sessile, broad-based antheridium having a free efferent tube; the fertile cells then commonly delimit another antheridium on one or both sides.Sometimes a fertile cell, especially one near the base of the appendage, gives rise not only to an antheridium but also to another cellular sterile or fertile branchlet.
By the time perithecial production begins (Fig. 18), the primary receptacular upgrowth, subtended by the enlarged and protruding cell lib, consists of a relatively massive, more or less vertical assemblage of about five progressively smaller cells and numerous small lateral cells.This complex of cells bears a relatively large number of sterile or antheridial appendages on its posterior surface (Fig. 18-19).
Soon after the primary receptacular upgrowth begins to develop, cell Ila' initiates the formation of the secondary receptacular upgrowth (sru) mentioned above.This cell has divided once in the individual shown in Fig. 10 and a small cell has been cut off distally on the outside.The first-formed cell then divides and forms a second cell (Fig. 11) and so on until 4-5 cells have been produced; each of these cells may cut off one or more cells to the right or left.The secondary upgrowth does not become vertically oriented as does the primary upgrowth.Instead, it is a more or less transverse group of cells subtended by the relatively large cell Ila' .As soon as they are delimited, cells of the secondary upgrowth give rise distally to usually fertile secondary appendages identical to those formed by cells of the primary upgrowth (Fig. 11,(18)(19)(20).A young secondary appendage is shown in Fig. 12, apd several mature or nearly mature appendages bearing successive ranks of antheridia are shown in Fig. 11,[18][19][20] At about the stage of development shown in Fig. 18, perithecial formation begins.The first perithecial initial appears to arise always from the upper end, either on the right or left side, of the original cell Ila' (Fig. 18-19).The side of the receptacle opposite the first perithecial initial bears secondary appendages.Additional perithecia usually arise from inner cells of the secondary upgrowth where they are flanked by secondary appendages.Though uncommon, perithecia also may arise from the primary receptacular upgrowth, especially from distal cells of the median row of cells (Fig. 22-24).The formation of perithecial initials always is progressive, never simultaneous.
Of 213 individuals bearing one or more mature or immature perithecia, 186 had perithecia arising only from the secondary upgrowth.Only one individual bore a perithecium on the primary upgrowth without any evidence of a perithecium having been initiated on the secondary upgrowth; 26 individuals bore immature or mature perithecia on both the primary and secondary upgrowths.Fourteen of the latter thalli bore one or more damaged or aborted perithecia on the secondary upgrowth-often the first-formed perithecium.
The number of perithecia initiated by a thallus varied from one to seven in the 213 individuals mentioned above; however, those bearing only one or two mature perithecia were most common.One hundred and seven individuals bore mature perithecia as follows: 82 with one; 19 with two; 3 with three; 1 with four; and 1 with seven.In the latter individual (Fig. 24) two of the perithecia arose from the primary receptacular upgrowth.
Soon after perithecial development begins (Fig. 14, 18), the stalk cell (VI), secondary stalk cell (VII), and basal cell m begin to elongate (Fig. 15,16,19) and rapidly carry the body of the perithecium together with the other two basal cells, nand n', upward and away from the level ofthe appendages (Fig. 20).At maturity, the greatly elongated cells VII and m constitute at least one third or more often one half of the total length of the perithecial stalk (Fig. 21,22,24).Basal cell m typically is slightly longer than the body of the perithecium proper (Fig. 17).
The trichogyne (tr) first appears as a budlike upgrowth from the tip of the trichophoric cell (tc) at a stage of development slightly later than that shown in Fig. 14.One to three simple or two-to three-times divaricate, septate branchlets then grow out from the more or less ovoid upgrowth, and each branchlet becomes bi-or trilobate at its apex (Fig. 15,19).None of the trichogynic septa become externally blackened.The trichogyne often persists at the apex of the young perithecium through at least the two wall-cell stage (Fig. 20).When it disappears, it leaves an inconspicuous basal remnant that persists on the surface of the peri thecial wall near the distal end of the third wall cell of the row of wall cells derived from basal cell m.
The mature perithecial body is relatively short compared to the elongate stalk ; it is broadest near the middle, nearly symmetrical, and  Individual showing two mature perithecia, one developed from the primary receptacular upgrowth, the other from the secondary receptacular upgrowth.This individual also bears, from the secondary upgrowth, an immature perithecium (projecting upward to the left [per]) and the broken base of the first-formed perithecium which had arisen from cell Ila' (see next figure), Xl60 (bar= 100 ~tm).-23.Enlargement of individual shown in Fig. 22 showing more clearly the origin of the two mature perithecia from the primary and secondary receptacular upgrowths (pru, sru).The upper end of the broken primary stalk cell of the primary perithecium formed on the secondary receptacular upgrowth is indicated by an arrow, X300 (bar= 50 ~tm).-24.Aged individual showing seven mature perithecia.The two at the left have arisen from the primary receptacular upgrowth; the five at the right (two with damaged or aborted perithecia [top]) have arisen from the secondary upgrowth, Xl25 (bar= 100 ~tm).-25.Ascospores showing the hyaline adhesive sheath and the hyaline endpiece terminating the longer of the two cells of the spores, X850 (bar= 20 ~tm).

@
® uniformly tapered to the blunt apex which is distinguished only by an inconspicuous median protuberance on the upper margin of each terminal cell (Fig. 17).There are four progressively smaller cells in each vertical row of outer wall cells.Cells of the lowermost tier are largest and up to 40% longer than those of the subbasal tier; cells of the terminal tier are smallest and slightly less than one half the length of those of the subterminal tier (Fig. I 7).There are four ascogenic cells.
The mature thallus is nearly uniformly yellow brown or orange brown, except for the basal cells of the appendages, the junctions between septa of the receptacle, and the posterior surface of cell lib which usually are darker.
Likewise, the transverse sep~a separating cell VI from cells VII and m and those near the base of the perithecium as well as the surface ofthe third tier of outer wall cells are more deeply pigmented than adjacent cell surfaces.

B. Supplemental Observations on Development of Presumably Allied Taxa
In the commentaries on Symplectromyces, Teratomyces, Sandersoniomyces, and Diplomyces to follow in this section, all references to my earlier study (Benjamin 1968a) on representatives of these genera will be indicated simply by the year date 1968a and the appropriate page number and/or figure number.

Symplectromyces vulgaris
Several stages of early development of S. vulgaris were described and illustrated in my earlier account of the species (l968a:8-10, Fig. 5A-F); however, there were certain errors in this treatment which need to be corrected together with some addenda.
My account of early stages of development of the receptacle and primary appendage given previously (1968a:8, 10, Fig. 5B-C) contains several errors of interpretation.In the stage of development shown in Fig. 27, the primary receptacle consists of three cells, I, II, and III, separated from the young three-celled primary appendage by the original spore septum (a) which in 1968 I wrongly interpreted as being the darkened septum above cell b.Contrary to my earlier thoughts (1968a:8, Fig. 5C), it is cell II, not cell III, that is divided by a more or less oblique cross wall into two cells, Ila and lib (Fig. 28), resulting in a primary receptacle of four superimposed cells.
The terminal cell of the three-celled primary appendage shown in Fig. 27 grows outward and upward past the spinose process which diverges near its base, forming a simple sterile prolongation which becomes septate in age.The cells immediately below also may give rise to short sterile branchlets (Fig. 28-31; l968a:Fig.5C, E-F).Two or three additional branchlets usually arise from basal cell b of the young primary appendage and the septa separating each of these branch lets from cell b become blackened (Fig. 2-31; 1968a:Fig.5E-F).Most intercalary cells of these branchlets are converted into simple antheridia (Fig. 30-31; l968a:Fig.5E-F).
While the primary appendage still is in early stages of development, cell III of the young receptacle begins to divide and cut off small cells around its periphery (Fig. 30-31; l968a:Fig.50-E).As the receptacle enlarges, cells I, IIa, and lib remain undivided, but the number of cells surrounding cell III continues to increase and from them are developed numerous simple or branched, sterile or antheridial, secondary appendages and one or more perithecial initials (1968a:Fig.5H).Secondary appendages typically have blackened septa separating the basal cell from the branch or branches above (1968a:Fig.5G): in age, proximal and distal cells tend to elongate and the antheridial cells may proliferate and give rise to short, divergent, sterile branchlets (1968a:Fig.5I).
Several stages of development of the perithecium of S. vulgaris were described and illustrated earlier (1968a: 11-12, Fig. 6).The trichogyne usually becomes several times branched as it matures, but only the septum separating the lowermost cell from the trichophoric cell becomes externally blackened (1968a:Fig.5H, 6C).This septum forms a remnant on the perithecial wall that appears always to lie near the lower margin of the terminal outer wall cell of the row of wall cells derived from basal cell n adjacent to the basal cell n' -derived row of outer wall cells.This remnant may even straddle the vertical septum separating these cells.The mature perithecial body together with the relatively large basal cells nand n' usually is nearly equal in length to the perithecial stalk (1968a:Fig.51).As in/.peyritschii, basal cell m is much longer than basal cells nand n' and its lower one half or more parallels the secondary stalk cell (VII) which subtends cells n and n'.Cells VII and m often contribute nearly one half of the total length of the stalk (1968a: Fig. 51) although the primary stalk cell (VI) may be longer in some individuals.Unlike I. peyritschii, however, basal cell m always is much shorter than the body of the perithecium.There are four or eight ascogenic cells.
Each vertical row of outer wall cells of the mature perithecium consists of four cells ( 1968a:Fig.51).Cells of the lowermost tier may be up to 20% longer than those of the subbasal tier and together these tiers constitute about four fifths of the total length of the body of the perithecium.Cells of the terminal tier are relatively small and inconspicuous, being one half to one third the length of cells of the subterminal tier.
Perithecia arise progressively, never simultaneously, on the receptacle.In his description of S. vulgaris, Thaxter (1908) reported one to three perithecia per individual, although in one of his figures (Thaxter 1908, Pl. L, Fig. 15) he depicted a specimen with four.I have observed up to four immature perithecia on developing specimens, but I have found that most commonly only one or two perithecia mature.In a sample of 203 adult specimens, the count of mature perithecia per thallus was as follows: 1 on 131 thalli; 2 on 69 thalli; and 3 on three thalli.

Teratomyces mirificus
The illustrations accompanying my brief discussion ofT.miri.ficus in 1968 did not include thalli in early stages of development.I have reexamined hosts from which material was taken earlier and have found a number of immature specimens several of which are figured here.Although these have not given as clear a picture of the earliest stages of growth of juvenile thalli as could be desired, they do suggest that development of this species is similar to that of S. vulgaris.
There is no evidence that the spore apex forms a persistent spine as in I. peyritschii and S. vulgaris.However, it proved difficult to remove very young individuals of T. miri.ficusintact, and the primary appendages of specimens observed were often more or less broken .The lower cell of the ascospore consistently displays a distinct subterminal constriction (Fig. 32; 1968a:Fig.4G).
I could not find a stage of development earlier than the one shown in Fig. 33 where the young receptacle consists of four superimposed cells separated from a four-celled primary appendage by a cross wall, a, believed on the basis of later development ofthe thallus, to represent the original cross wall of the ascospore.Even at this stage, the lower two cells of the young receptacle above and below the cross wall separating them have in part become nearly obscured by darkening and thickening of their outer walls; this is characteristic of most thalli of this species.Although the origin of the middle two cells of the receptacle shown in Fig. 33 from the division of a previously formed cell II, as in S. vulgaris, could not be determined, I have designated the four cells I, Ila, lib, and III at this stage of development for sake of consistency.
In the young individual shown in Fig. 34, the tip of the primary appendage is missing, but a single branchlet is developing from its basal cell.Figure 35 shows another young thallus in a slightly more advanced stage of development.The primary appendage (pa) is broken and consists only of the original basal cell and a single three-celled branchlet with a pointed tip.Cell III has been divided into two cells by a vertical cross wall, and the daughter cell has given rise to the first secondary appendage.Further proliferation of cell III and its progeny and the continued formation of secondary appendages are shown in Fig. 36-38.Cells I, Ila, and lib of the receptacle do not divide further, but there is rapid enlargement and elongation of the receptacle, especially cell lib, along with an increasing number of appendiculate cells and secondary appendages ( 1968a:Fig.4A-B).
Secondary appendages, depending on their age, vary from simple to highly branched ( 1968a:Fig.4B-F).They may consist only of two or three small superimposed cells, with the terminal cell more or less curved and beaklike.Or, more commonly, they consist of a relatively large elongate basal cell bearing a unilateral series of free, elongate, sessile phialides and several lateral or terminal sterile or fertile branchlets bearing phialides or beaklike cells.Cross walls typically are blackened externally.
Perithecia arise successively from inner cells ofthe complex of cells arising from cell III of the receptacle ( 1968a:Fig.4A).As the receptacle enlarges and the number of appendiculate cells increases, a relatively large conical cell may be evident in the center of this mass of cells and probably represents the original cell III (1968a:Fig.4B).Whether or not perithecial initials arise from this cell could not be determined for sure; if not, they certainly arise from cells closely associated with it.The trichogyne of T. mirificus is more or less highly branched and septate (1968a:Fig.4H; Thaxter 1896, Pl.II, Fig. 1) with the basal septum and those of the lowermost branches externally blackened.The basal septum remains as a conspicuous trichogynic remnant on the outer wall of the perithecium, usually near the upper end of the juncture between the subterminal outer wall cells of the m-and n-basal cellderived rows of wall cells.There are two ascogenic cells.
Each vertical row of outer wall cells of the mature perithecium of T. mirificus consists of four cells ( 1968a:Fig.4C).The basal and subbasal tiers are nearly equal in length and together they constitute about 85% of the total length of the body of the perithecium.The very small terminal outer wall cells are about one half the length of the subterminal cells.Basal cell m and the secondary stalk cell (VII) usually form a relatively small part of the perithecial stalk, and their combined length together with basal cells n and n' is about equal to that of a basal outer wall cell (1968a:Fig.4C).The primary stalk cell (VI) may vary greatly in length, commonly being two times and often three to four times the length of the perithecium (1968a:Fig.4C; Thaxter 1896:Pl.X, Fig. 4, 193l:Pl.XLIII, Fig. 4).
Like Thaxter (1896), I have observed only one to three mature perithecia on adult specimens of T. mirificus, although I have seen as many as four mature and immature perithecia on some thalli.The number of mature perithecia per individual in a sample of 94 specimens examined was as follows: 1 on 7 5 thalli; 2 on 1 7 thalli; and 3 on two thalli.
In 1968, I reported a taxon thought to be a Symplectromyces found on Philo nth us Curtis spp.collected in Colorado (RKB 2334), New Mexico (RKB 2339), and California (RKB 2351(RKB , 2352)).Further study of these collections shows that the taxon has ascospores with a subterminal constriction like that of T. mirificus.Also, its perithecium has only two ascogenic cells, and its secondary appendages resemble those of other Teratomyces species in having relatively large basal cells with blackened, stepwise septa delimiting the lateral branchlets.However, beaklike cells are lacking, and its antheridia are formed like those of S. vulgaris, i.e., they are formed directly from intercalary cells of fertile branchlets.This taxon may be distinct, but it is tentatively being referred to T. philonthi Thaxt.(1900).These observations support Balazuc's (1974) contention that Thaxter's (1908) report of S. vulgaris on a Philonthus from Hungary probably was based on an erroneous host determination.

Sandersoniomyces divaricatus
Several additional drawings of immature specimens of S. divaricatus are added here to supplement those accompanying my description of the species in 1968.The tip of the lower cell of the ascospore is without special modification (Fig. 39).Reexamination of the youngest individual illustrated earlier ( 1968a:Fig. 1 B) has revealed what appears to be the spinose remnant of the spore apex on the upper cell of the right branchlet of the primary appendage (pa) (Fig. 41 ).However, this branchlet apparently is easily broken or is readily deciduous, and the consistent presence of a spinose appendage cell could not be determined.No stage of receptacular development earlier than that shown in Fig. 40 has been encountered.As in the other species being discussed in this paper, the four cells ofthe young primary receptacle are numbered I, Ila, lib, and III from the base.
The primary appendage (pa) of the individual shown in Fig. 40 is broken and only the basal cell and part of the next cell are intact above what is believed to be the original spore septum (a).The primary appendage ofthe specimen shown in Fig. 41 is intact and a three-celled branchlet already has arisen from the basal cell next to the spinose primary branchlet.The secondary branchlet is terminated by a beaklike cell with a minutely rounded apex which is characteristic ofthe pointed terminal cells of sterile branchlets in this species (Fig. 43-44; l968a:Fig.1 B-F; 2C-G).As thalli mature, primary appendages were not observed on any maturing individuals studied.Some branches may be deciduous, for basal cells of primary appendages usually bear one or more circular scars left by the basal septa of detached branchlets (Fig. 43; 1968a:Fig.1 E; 2B).
As the receptacle enlarges, cell III gives rise to a pair of one-celled outgrowths on one side below the primary appendage, and from these cells axes of superposed cells develop which give rise to secondary appendages and perithecia (1968a:Fig. 1 E-F, 2A-B).Figure 42 shows a young individual in which the first of these outgrowths has developed.A slightly later stage is shown in Fig. 43.This individual is seen from the rear and shows the median primary appendage and, on the far side, the first secondary appendages arising from the pair of cells formed by cell III.A somewhat older individual as seen from the front is shown in Fig. 44 where the third cells of the succession of cells forming the fertile cellular axes are arising from the second, already appendiculate, cells.Stages of development comparable to those shown in Fig. 43  Each cell of a fertile axis bears one secondary appendage distally on its posterior side ( 1968a:Fig.2A-B).Secondary appendages are more or less uniform in structure, consisting of a relatively large basal cell subtending externally a three-celled, sterile branchlet with a terminal beaklike cell having a minutely rounded tip and internally a several times branched branchlet the distal cells ofwhich are converted into antheridial cells (1968a:Fig.2G).
Proximal cells of primary cellular axes may give rise to additional axes or to perithecial initials ( 1968a:Fig. 1 F, 2A-B).Axis cells distal to the one giving rise to the first perithecial initial may give rise to one or more initials; however, unless the first-formed perithecium on an axis is damaged or aborts the others apparently do not develop.Whether or not this is the usual situation in this species remains to be determined, for only 12 of the 9 5 individuals examined thus far bore mature perithecia.Of these, 11 bore one perithecium, the other bore two.
The mature trichogyne of S. divaricatus is more or less highly branched and septa separating its lower cells are externally blackened ( 1968a:Fig. 1 G).Its lower part may remain attached at or near the perithecial tip until at least the first three outer wall cells have been delimited.When it finally sloughs away, its basal septum forms a subterminal scar on the perithecial wall at or near the juncture of the two terminal outer wall cells of the m-and adjacent n-basal cell-derived rows of wall cells.There are two ascogenic cells.
Outer wall cells of the perithecium are disposed in four vertical rows of four cells each (1968a:Fig.lA-B).Cells of the basal tier are only slightly longer than those of the subbasal tier on the average, and together these tiers constitute 85% or slightly more of the total length of the body of the perithecium.Cells of the terminal tier are only slightly shorter than those of the subterminal tier.The combined length of the basal cells and secondary stalk cell is considerably less than that of cells of the lower tier of outer wall cells ( 1968a:Fig.2A-B).The primary stalk cell in all but one of the 12 mature individuals examined was about equal in length to the combination of basal cells and secondary stalk cell (1968a:Fig.lA-B).In the one exception, this cell was about twice as long as any of the others, and only with a larger sample of mature specimens could one determine the degree of variability in the dimensions of this cell which in species of many genera of Laboulbeniales often ranges most widely in length.

Diplomyces actobianus
Early stages of development of D. actobianus prior to perithecial initiation were not included in my 1968 paper.Two are added here, but the earliest stages of growth following spore germination still have not been encountered in the material available for study.Ascospores are relatively slender (Fig. 45) but are not otherwise different from those usually found in Laboulbeniales.Thaxter (1896:Pl.X, Fig. 20) figured an immature individual younger than any found by me.He shows a four-celled receptacle in which the cross wall separating the middle two cells is oblique, suggesting that the subbasal cell of the young receptacle divides and forms two cells in a manner similar to that found in I. peyritschii and Symplectromyces vulgaris (Fig. 5, 28).
Again, I have numbered the cells of the young receptacle I, Ila, lib, and III.
I have not observed a young primary appendage of D. actobianus at a stage of development comparable to that shown for Sandersoniomyces divaricatus in Fig. 41.The two terminal branches on the young individual figured by Thaxter ( 1896:Pl.X, Fig. 20) appear to be attached to cell III of the receptacle which is shown undivided.One of these branches surely is the primary appendage.The other could be a branch of this appendage, incorrectly depicted, or it could be the first branch of a secondary appendage arising from a cell cut off on the far side of cell III.The primary appendage of D. actobianus soon is obscured by development of the two posterior upgrowths characteristic of this species and by secondary appendages and perithecia (Fig. 47; 1968a:Fig.3A-C).However, observations on some 30 immature specimens indicates that the primary appendage of D. actobianus is similar to that of S. divaricatus and consists of a relatively large basal cell bearing several branchlets, some of which form beaklike terminal cells.Whether or not cells of some branchlets form antheridia still is undetermined.
As inS.divaricatus, cell III of the young receptacle of D. actobianus forms a pair of anterior cellular upgrowths on one side below the primary appendage.These cellular upgrowths give rise first on the outside to single, secondary, antheridia-bearing appendages (Fig. 46; 1968a:Fig.30) followed soon on the inside by cellular upgrowths (Fig. 4 7) the successive cells of which bear single antheridial secondary appendages (1968a:Fig.3E) and, later, secondary perithecial initials (1968a:Fig.3F).Lastly, the paired cells give rise, anteriorly and distally, to single primary perithecial initials (1968a: Fig. 3A).Soon after the paired anterior cells begin forming the structures listed above, cell III also initiates formation, posteriorly and distally, of the pair of sterile, two-celled upgrowths characteristic of the species (Fig. 47; 1968a:Fig.3B).
The first-formed secondary appendages arising on the outside ofthe paired anterior cells of the receptacle consist of relatively large basal cells bearing one or more three-celled, beaked, sterile branchlets and several fertile branchlets the cells of which are converted into antheridia (1968a:Fig.3A,  D).In age, the appendage may become distally more or less sympodially branched, each branch bearing several sterile and fertile branchlets, and the basal cell may become somewhat elongate with a unilateral series of sterile or fertile branchlets.These appendages never approach the inner cellular axes in length or number of branches and they do not give rise to perithecial initials.The secondary appendages formed on the elongate inner branches are comparatively simple, most commonly consisting of a basal cell bearing externally a beaked three-celled branchlet and one or several antheridial branchlets ( 1968a:Fig.3B, C, E).Cross walls separating cells of appendages typically are blackened.
Despite the more or less symmetrical development of the receptacle and its appendages and sterile upgrowths, the perithecia of D. actobianus develop and mature progressively.The first-formed or primary perithecia arise from first one then the other of the anterior cellular outgrowths of cell III ( 1968a: Fig. 3A).Both primary perithecia may eventually mature and the result is a symmetrical thallus like the one shown in Fig. 3C of my 1968 paper.Or as shown in Fig. 3B of the same work only one perithecium may mature.Additional perithecial initials arise not from the paired anterior cells of the receptacle but from the basal and later succeeding cells of the inner cellular upgrowths formed by the paired cells ( 1968a:Fig.3F).This occasionally results in a thallus bearing more than two mature perithecia, i.e., two primary and one or perhaps two secondary perithecia.I have not seen mature secondary perithecia on cells other than the basal cells of the cellular upgrowths.Only 23 of about 50 specimens studied bore mature perithecia.Eight had only one primary perithecium; 13 had two primary perithecia; and 2 had three perithecia, two primary and one secondary.One of the latter specimens bore an additional immature secondary perithecium that appeared normal and probably would have matured eventually.
The trichogyne of D. actobianus is at first simple (1968a:Fig.3A) but becomes more or less branched with the proximal cells slightly constricted at their blackened septa.The basal cell and often several cells of the lowermost branch or branches of the trichogyne usually remain broadly attached to the outer surface of the perithecium near or across the junction of the two terminal outer wall cells of the rows of wall cells derived from basal cell n (1968a:Fig.3C).There are two ascogenic cells.
Perithecial outer wall cells are disposed in four vertical rows of four cells each, with those of the lower two tiers nearly equal in length and together comprising ca.80% of the length of the body of the perithecium (1968a:Fig.3B, C).Cells of the subterminal tier are only slightly longer than those of the terminal tier.The primary stalk cell appears never to elongate greatly and typically is equal to or only slightly longer than the combined lengths of the secondary stalk cell and basal cells.The stalk cells and basal cells together constitute about one third of the total length of the perithecium.

A. Hosts
Species of Diplomyces, Sandersoniomyces, Symplectromyces, Teratomyces, and Idiomyces are known only on beetles belonging to the Staphylinidae.This huge family of insects, on a worldwide basis, was divided into 18 subfamilies by Blackwelder (1952) and of these eight have been confirmed as including hosts of one or more species of perhaps 53 genera of Laboul-benia1es (55 with the addition of two segregate genera of Misgomyces Thaxt. [ 1900] to be recognized by Tavares in an as yet unpublished major revision of these fungi).Modifications of Blackwelder's treatment of the subfamilies Oxytelinae (Arnett 1961;Frank 1982) and Staphylininae (Frank 1982) increase the number of host subfamilies to 12. Frank's (1982) definitive survey of staphylinid hosts of Labou1beniales serves as the basis of the following summary of the genera of Laboulbeniales occurring on Staphylinidae.
All species of 34 genera of Laboulbeniales are known only on Staphylinidae, whereas 19 genera are known not only on Staphylinidae but also on other families of arthropods, mostly true insects.Of the genera found only on Staphylinidae, one or more species of 29 are limited to insects belonging to only one of nine subfamilies: Omaliinae: Balazucia Benjamin ( 1968b ); Osoriinae: Osoriomyces Terada (1981)  (1894), is found on three subfamilies, i.e., Aleocharinae (1 of 34 species), Staphylininae, and Xantholininae.Monoicomyces Thaxt.(1900) has been found on members of five subfamilies, i.e., Aleocharinae, Oxytelinae, Osoriinae, Tachyporinae, and Xantholininae.There is an apparent high degree of specificity among genera and species of Laboulbeniales parasitizing Staphylinidae and Frank's ( 1982) study should be consulted in this regard.
Members of subfamily Staphylininae sensu Frank ( 1982) serve as hosts for all species of Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces.Blackwelder's (1952) Staphylininae included three tribes, Xan- tholinini, Staphylinini, and Quediini.Arnett (1961) retained only the Xantholinini and Staphylinini and placed the Quediini in the latter tribe as one offive subtribes, only three of which, Staphylinina, Oxypina, and Quediina, occur in the United States.Frank (1982) narrowed the concept of Staphylininae by recognizing a separate subfamily, Xantholininae, for the Xantholinini.His concept is the one adopted here.Insects of only two of the five subtribes of Staphylinini, Oxypina and Quediina, host all of the above genera of Laboulbeniales.Among the Oxypina two genera are hosts: Erichsonius Fauvel (Teratomyces actobii Thaxt.[1894], T. brevicaulis Thaxt. [1894], Diplomyces actobianus, and D. clavifer Rossi & Cesari-Rossi [1978]) and Gabrius Curtis (as Philonthus) (T.philonthi).Among the Quediina three genera are hosts: Acylophorus Nordmann (T.mirificus); Atanygnathus Jacobson (D. atanygnathi Thaxt. [1931]  Idiomyces peyritschii is found only on species of De/easter.This genus is included by Arnett ( 1961) in subtribe Coprophilina of the Oxytelini, one of eight tribes of the subfamily Oxytelinae.The concept of this subfamily as treated by Arnett and Frank (1982) is more restricted than that of Blackwelder (1952) in that the latter's Oxytelinae is divided into five subfamilies of which four, Oxytelinae, Piestinae, Omaliinae, and Osoriinae, are known to include hosts of Laboulbeniales.This sytem essentially follows that of Crowson ( 19 55) who places De/easter in a separate tribe, Deleasterini, as does Hatch (1957).The Coprophilina of Blackwelder and Arnett includes 12 genera ofwhich, in addition to De/easter, only Manda Blackwelder (syn.Acrognathus Erichson) has a species, M. mandibularis (Gyllenhal), known to be a host of one of the Laboulbeniales, Peyritschiella pro tea Thaxt.( 1900), which also has been reported on species of Anotylus Thomson, Bledius Leach, Oxytelus Grav., and Planeustomus Jac. de Val ofsubtribe Oxytelina of the Oxytelini.As a genus, Peyritschiella Thaxt.( 1890) infects a broader spectrum of hosts than Idiomyces, and its species are known not only on other groups ofStaphylinidae (suborder Polyphaga) but on Carabidae (suborder Adephaga) as well.
Among the Staphylinidae, Crowson (1955) regards the Staphylininae (including the Xantholininae as a tribe, Xantholinini) together with another subfamily, Paederinae, to be phylogenetically the most highly evolved groups in the family and not closely connected by transitional forms with the oth er subfamilies.The larvae ofOxytelinae appear to be closely related to those of Aleocharinae and bear no special resemblance to those of Staphylininae s. 1. (Boving and Craighead 19 31 ).Paederinae and Staphylininae host 13 genera of Laboulbeniales, listed above, not found on any other group of insects, and the Oxytelinae are host to four such genera.Of 19 genera known not only on Staphylinidae but also on other families of insects as well, only two, Laboulbenia and Peyritschiella, have been confirmed on both Oxytelinae and Staphylininae.With regard to their laboulbeniaceous parasites, Oxytelinae and Staphylininae appear to have little in common, reflecting perhaps early divergence in the evolutionary history of these insect subfamilies and their parasites.
One must be cautious, however, in making taxonomic judgments based on host range in the Laboulbeniales.There are some genera, perhaps representing the most ancient lineages, that have developed a remarkable capacity for exploiting different host groups, e.g., Dimeromyces Thaxt.( 1896), known on Acarina, Coleoptera, Dermaptera, Diptera, Hymenoptera, Isoptera, and Orthoptera; Laboulbenia, known on Acarina, Blattaria, Coleoptera, Diptera, Hemiptera, Hymenoptera, Isoptera, and Orthoptera; and Rickia Cavara (1899), known on Acarina, Blattaria, Coleoptera, Diplopoda, Hymenoptera, and Orthoptera.One also must be aware of the possible spread of these parasites to unrelated host groups which happen to be closely associated in a very restricted environment.This undoubtedly has led to the occurrence of Trenomyces Chatton & Picard (1908) on such diverse insects as members of the Haematopinidae (Anoplura) (Benjamin unpubl.),Hippoboscidae (Diptera) (Bequaert 1953(Bequaert , 1957)), and Laemobothriidae, Menoponidae, and Philopteridae (Mallophaga) (Chatton and Picard 1908;Eichler 1939Eichler , 1942Eichler , 1951;;Thaxter 1912bThaxter , 1924) ) which are specialized ectoparasites of birds and mammals.Finally, survival and subsequent development of a fungus following its fortuitous transfer from one host to a wholly unrelated host in any environment must account for marked disjunctions in host range occasionally observed (Benjamin 1967), and this kind of dispersal probably has played a major role in the diversification ofLaboulbeniales, at both the genus and species levels, on related as well an unrelated host groups.However, as in the Staphylinidae, most genera of Laboulbeniales are restricted to a single taxon or a few related taxa of hosts, e.g., Coreomyces Thaxt.(1902) on Corixidae (Hemiptera), Herpomyces Thaxt.(1902) on Blattaria, Tettigomyces Thaxt.(1915) on Gryllotalpidae (Orthoptera), and Eucantharomyces Thaxt. (189 5)on Carabidae (Coleoptera); one genus, Ilytheomyces Thaxt. ( 191 7),is found only on a single genus, Ilythea Curtis, of the Ephydridae (Diptera).All of these genera are large enough in number of known species and are sufficiently widespread throughout the world to indicate that they do, indeed, have fixed host requirements.
The high degree of host specificity exhibited by so many genera of La-boulbeniales occurring on Staphylinidae suggests that closely related genera may be expected to occur on closely related rather than on distantly related hosts, even within the same host family.With 14 taxa restricted to five allied genera of insects, Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces exhibit a degree of host specificity suggesting that their presumed near relatives would also be found on similar hosts.

B. Ascospores
Ascospores of Laboulbeniales display a remarkable uniformity in shape and structure, being in all instances colorless, elongate, narrow, sometimes slightly rounded but mostly acute at both ends, and once septate.All are surrounded by a hyaline adhesive sheath that usually is more or less expanded at the foot-forming end of the spore.The position of the cross wall usually is off center with the longer cell directed towards the perithecial ostiole prior to spore discharge.In a few genera, i.e., Ceratomyces Thaxt.( 1892), Coreomyces, Rhynchophoromyces Thaxt. (1908), and Zodiomyces Thaxt.( 1891), the shorter of the two cells is uppermost prior to spore discharge (Thaxter 1896(Thaxter , 1908) ) and eventually forms the foot.The cross wall typically is transverse, although in some genera it may be slightly oblique, e.g., Dioicomyces Thaxt.(1901,1908,1931).The adhesive sheath, except where enlarged distally at one end, usually is continuous around the spore, although in species of Autoicomyces Thaxt.( 1908) it may be constricted at the spore septum where it appears to be attached to the spore wall (Thaxter 1896(Thaxter , 1908)).In some dioecious genera, e.g., Dioicomyces, there is a marked difference in size of male and female spores, the latter being much the larger (Thaxter 1908;Benjamin 1970).In the first two parts of his monograph, Thaxter (1896Thaxter ( , 1908) ) often figured ascospores and pointed out deviations from the norm as mentioned above.However, in the last three parts of his illustrated work (Thaxter 1924(Thaxter , 1926, 19 31) , 19 31) ascospores rarely were figured separately and minor ascospore anomalies, if such existed, beyond those he had pointed out earlier were not discussed.Because in most genera of Laboulbeniales there is so little variation in the morphology of ascospores, other than size, spores have been little emphasized in the taxonomy of these fungi above the species level.
Ascospores of Diplomyces, Sandersoniomyces, and Symplectromyces (Fig. 26, 39, 45) lack any special modifications to set them apart from the spores of most other Laboulbeniales, at least as they have been described in the literature.In species of these genera the foot-forming cell is the longer of the two cells comprising the spore body; the spore septum is transverse; the cellular apices are acute; and the thin hyaline sheath is more or less expanded around the lower one fourth to one third of the spore.Ascospores of the known species of Diplomyces, D. actobianus, D. atanygnathus, and D. clavifer, based on study of specimens in my collection, differ from those of species of the other genera being considered here in their slenderness, being about one half as wide.
The distinctive modifications of the tip of the foot-forming cell of the ascospores of I. peyritschii (Fig. 1, 25) and T. mirificus (Fig. 32) have not, as far as I am aware, been reported for any other genera of Laboulbeniales.A conspicuous subterminal constriction of the basal cell like that of T. mirificus also is characteristic of the taxon on Philonthus spp. earlier identified with T. philonthi.I have a number of collections referable to T. actobii and T. brevicaulis and have not found a similar obvious constriction of the lower cell of their ascospores.This phenomenon and its possible significance in the taxonomy of Teratomyces should be taken into account in future studies of the genus.
The hyaline end-piece terminating the truncate foot-forming cell of the ascospore of I. peyritschii apparently is unique in the Laboulbeniales.If this structure were correlated with early stages of spore germination, then similar structures surely would have been reported in other species.For the same reason, it probably is not associated with the formation of the often-observed more or less hyaline tip of the foot (Fig. 2-6, 27-29) in juvenile and adult individuals of other species (Benjamin 1968b, 1973, andunpublished).

C. Receptacle
Divisions of the lower segment of the ascospore give rise to the receptacle comprised, depending on the taxon, of a small, fixed number of cells, i.e., a determinate receptacle, or of a variable, often large number of cells, i.e., an indeterminate receptacle (Thaxter 1908).In the simplest type of determinate receptacle, the primary receptacle derived from the lower spore segment remains three celled; the lower cell (I) forms the foot, the suprabasal cell (II) gives rise to the perithecial branch, and the upper cell (Ill) subtends the primary appendage derived from the upper spore segment, e.g., Stigmatomyces (Thaxter 1908(Thaxter , 1931)), Haplomyces (Thaxter 1896) and their allies.Exceptions are found in Amorphomyces and Rhizopodomyces Thaxt.(1931) in which a cell III is not separated from cell II (Tavares 1970;Benjamin 1979).Determinate receptacles having more than three cells are not common.Thaxter (1908) made a case of Ceratomyces and Autoicomyces in which the primary receptacle consists of only three or four cells below the perithecium and appendage.The receptacle of most species of Laboulbenia (Thaxter 1896(Thaxter , 1908) ) consists of only five cells, designated by Thaxter with Roman numerals I to V. Cells I, II, and III correspond to those ofthe simple determinate receptacle as defined above; extra divisions of cell III give rise to cells IV and V which commonly parallel or are attached to the stalk cells of the perithecium derived from cell II.Thaxter used numerals VI and VII to designate the primary stalk cell and secondary stalk cells of the perithecium of Laboulbenia and these numbers now are used, following Tavares (1970Tavares ( , 1973)), for describing these perithecial cells in all members of the Laboulbeniales.In many species of Laboulbenia, especially some of those on Chrysomelidae (Coleoptera) and Diptera, cell III does not divide and the receptacle is like that of the simplest determinate type.In other species, however, e.g., L. pro!iferans Thaxt.( 1908( ), L. variabilis Thaxt. (1908)), etc., cells IV and/ or V divide and form several additional cells and the receptacle becomes more or less indeterminate.
Receptacles of Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces are indeterminate, but secondary development, including formation of perithecia and secondary appendages, is confined to a complex of cells derived from what can be interpreted as cell III of the primary receptacle, the lower cells being determinate in number and retaining their identity in the mature thallus.A fourth cell is added to the young primary receptacle by division of cell II (Fig. 28), but in none of these genera does a perithecium arise from either cell derived from cell II.Perithecial development from cell II is the rule in all genera having simple, determinate, twoor three-celled receptacles.Thaxter (1908Thaxter ( , 1931) ) regarded Symplectromyces as very closely related to Teratomyces.He originally had placed S. vulgaris in that genus (Thaxter 1900) but later (Thaxter 1908) separated it from Teratomyces because of its intercalary antheridia (compare Benjamin 1968a:Fig.40-F and 5G).The origin and arrangement of secondary appendages are similar in both genera, but, as noted by Thaxter (1896), the arrangement of appendages of Teratomyces spp.does not have the same appearance from all sides (Benjamin  1968a:Fig.4A-C) as it does in S. vulgaris (Benjamin 1968a:Fig.5H-I; Thaxter 1908:Pl.L, Fig. 14-16).There is in Teratomyces a hint of the kind of asymmetrical development ofthe upper receptacle that is so pronounced in Diplomyces (Fig. 46, 47) and Sandersoniomyces (Fig. 42-44).
Receptacular development is similar in Diplomyces and Sandersoniomyces.In species of both genera, cell III of the primary receptacle cuts off a pair of anterior cells which in Diplomyces may give rise to perithecia, secondary appendages, and secondary cellular axes bearing both perithecia and secondary appendages (Benjamin 1968a:Fig.3A-C); in S. divaricatus these cells give rise only to secondary cellular axes bearing perithecia and secondary appendages (Benjamin 1968a:Fig.2A-B).In both genera the posterior cell subtends a small, sterile primary appendage.In Sandersoniomyces this cell undergoes no further development other than enlargement (Benjamin 1968a:Fig.2B).In D. actobianus (Benjamin 1968a:Fig.3B) and D. atanygnathus (Thaxter 1931:Pl.XLIII, Fig. 1) the cell gives rise to a pair of sterile upgrowths, whereas D. clavifer (Rossi and Cesari-Rossi 1978:Fig. 3-4) develops only one such upgrowth.The basal cell of D. clavifer also is distinguished by the formation of a single, large, clavate, externally blackened upgrowth, of specific importance only, similar to those found on a few species of several other genera, e.g., Corethromyces (Thaxter 1908(Thaxter , 1931)), Meionomyces Thaxt. (1931), and Misgomyces (Thaxter 1931).
The receptacle of Idiomyces peyritschii is different in several ways from that of Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces although in the three-and four-celled stages of development of the young receptacle there are similarities (compare Fig. 4-5, 27-28, and Thaxter 1896: Pl. X, Fig. 20) which suggest that these genera may be related.The primary receptacle of I. peyritschii consists of five rather than four cells, with three cells (Fig. 4, 5, 7) rather than two being derived from cell II (Fig. 27,28).In /. peyritschii as in the other four genera, cell III proliferates (Fig. 8-11) and forms a cellular complex that gives rise to secondary appendages .Although perithecia occasionally arise from these cells in /. peyritschii , their production appears to be suppressed in favor of their development, along with secondary appendages, from cells of a second complex of fertile cells derived from cell II of the receptacle (Fig. 10,11,(18)(19)(20)(21).This might be interpreted as suggesting that phylogenetically I. peyritschii is more closely linked to an ancestor having a determinate receptacle with cell 11-derived perithecia than are species of Diplomyces, Sandersoniomyces, Symplectromyces, and Teratomyces.
The occasional formation ofperithecia by cells of the primary receptacular upgrowth in I. peyritschii may be in response to damage to or abortion of perithecia arising normally from cells of the secondary receptacular upgrowth.As noted earlier, 14 of 26 thalli bearing perithecia on the primary upgrowth showed clear evidence of damaged or aborted perithecia-often the first-formed-on the secondary upgrowth (Fig. 22,23).In the specimen shown in Fig. 24, the two functional perithecia borne on the primary upgrowth are accompanied by two aborted perithecia on the secondary upgrowth.This may be only coincidental; on the other hand it may be a teratological phenomenon similar to others frequently encountered in the Laboulbeniales (see Balazuc 1976 and references therein).

D. Appendages and Antheridia
One or several appendages are found on all Laboulbeniales, although these structures may not be apparent on all mature specimens, and especially on the males of some dioecious species.In his general discussion of appendages, Thaxter (1896Thaxter ( , 1908) ) distinguished two kinds, primary and secondary.The first is defined as having its origin from the upper spore segment of the ascospore, whereas the latter has some other origin, i.e., from cells of the receptacle or, in a few genera, from the perithecium.
Primary appendages of Diplomyces, Sandersoniomyces, Symplectromyces, Teratomyces, and Idiomyces all are more or less branched and multicellular.Some branches of the primary appendage of Symplectromyces vulgaris definitely form seriate antheridia (Fig. 30-31) like those of the secondary appendages of this species (Benjamin 1968a:Fig.5G).Antheridia were not observed on primary appendages of species of the other genera, and I suspect that they are sterile.However, this still must be confirmed or denied for species of Diplomyces, Sandersoniomyces, and Teratomyces.The absence of antheridia on primary appendages of any of the intact juveniles of /. peyritschii that I studied  indicates that this appendage in this species is indeed sterile.
Well-developed, multicellular, usually branched, fertile secondary appendages are characteristic of the above five genera.The relationship of these appendages to the primary appendage and/or secondary receptacles or secondary cellular axes of these taxa already has been discussed.A close relationship between D. actobianus, D. clavifer, and S. divaricatus is indicated by the similarity of their secondary appendages (compare Fig. 2D-F and3D in Benjamin 1968a andRossi andCesari-Rossi 1978, Fig. 3-4) and receptacles.In all these species, this appendage consists of a single basal cell bearing, externally, a usually three-celled, sterile, beaked branchlet and, internally, several branchlets forming series of intercalary antheridia resembling those of the appendages of Symplectromyces vulgaris (Benjamin 1968a: Fig. 5G).The elongate, simple or branched appendages of S. vulgaris, however, never bear beaked sterile branchlets (Benjamin 1968a:Fig. 5H-I).In this connection it should be noted that the secondary appendages of D. atanygnathus, as described by Thaxter (1931: 173, Pl. XLIII, Fig. 1), are not like those ofthe other two species of this genus and would seem to resemble more closely those of Symplectromyces.This also is suggested by specimens referable to D. atanygnathi from the Philippines in my collection (RKB 799).In its perithecial and ascospore characteristics and furcillate receptacular upgrowths, however, D. atanygnathi more closely resembles the other species of Diplomyces.
Secondary appendages of T. mirificus (Benjamin 1968a:Fig.4D-F) become sympodially branched and bear both free antheridia (phialides) and sterile branchlets with beaked terminal cells.Secondary appendages of all other species of Teratomyces (Thaxter 1896(Thaxter , 1908(Thaxter , 1931) ) are similar in their construction except perhaps for T. philonthi (Thaxter 1908).Thaxter's description of this species does not mention and his figures of it do not show phialidelike anteridia, nor does the photograph of it published recently by Rossi and Cesari-Rossi (1978:Fig. 9).Instead, the appendages as Thaxter depicts them and as they are shown by the Rossi photograph appear to be separated into short cells many of which are separated by diagonal, darkened septa.This construction is like that of the appendages of specimens in the series of collections on Philonthus in my collection which in a previous section of this paper was tentatively referred to T. philonthi.This taxon lacks free phialides; instead the antheridia develop directly from intercalary cells as in Symplectromyces.Teratomyces philonthi may represent a form linking Symplectromyces and other Teratomyces spp.Likewise, D. atanygnathi may be a link between Symplectromyces and other species of Diplomyces.
Secondary appendages of I. peyritschii may be branched or unbranched.They often have one or more elongate, cellular branchlets arising from basal and distal cells with each cell in between delimiting two or three simple, sessile, adnate antheridia placed side by side and separated from the mother cell by broad septa.The manner in which antheridia are formed in this species recalls that of Stigmatomyces (Thaxter 1908(Thaxter , 1931)), and on this basis alone Thaxter (1896) suggested near relationship between Idiomyces and Stigmatomyces despite the marked differences in their receptacles and the origin of their appendages, i.e., primary in Stigmatomyces vs. secondary in Idiomyces.In 1908, Thaxter speculated that Idiomyces might be a transitional type between Stigmatomyces and Teratomyces and its allies.On the basis of some similarities in their receptacles and the positioning of secondary appendages and perithecia, Idiomyces may be distantly related to Teratomyces.Symplectromyces, Diplomyces, and Sandersoniomyces, but a relationship between these five genera and Stigmatomyces must at best be remote.

E. Perithecia
Perithecial development in the five genera discussed in this paper is like that of all members of suborder Laboulbeniineae (Thaxter 1908) in which the walls of the perithecium grow up and around the carpogonium which is derived from a preexisting initial (Tavares 1977(Tavares , 1980) ) in contrast to members of suborder Herpomycetineae (Tavares 1981) in which the first wall cells are developed prior to formation of the carpogonial initial which grows upward between the surrounding wall cell rows.Perithecial development has been described for a number of genera ofLaboulbeniineae, e.g., Balazucia (Benjamin 1968b), Euceratomyces (Tavares 1980), Filariomyces (Shanor 1952), Laboulbenia (Benjamin and Shanor 1950;Thaxter 1896), Peyritschiella (Thaxter 1896), Rhizopodomyces (Benjamin 1979), Stigmatomyces (Thaxter 1896), and Symplectromyces (Benjamin 1968a).
Despite their often varying length and spatial separation from the asciger-ous part, i.e., body of the perithecium, the primary and secondary stalk cells, VI and VII respectively, are regarded as integral parts of the perithecium as a whole (Thaxter 1896).The length of these cells may vary depending on the species and even within a given species.In many Laboulbeniales the stalk cells are small and relatively inconspicuous, whereas in others theyespecially cell VI-may elongate and form a well-defined pedicel which elevates the body some distance above the level of the receptacle.
The three perithecial basal cells, m, n, and n', give rise to the four rows ofperithecial wall cells and typically enclose the lower part of the ascigerous cavity in most Laboulbeniales.In a few genera the basal cells and stalk cells become disorganized, in part, as the perithecium matures, and the ascigerous cavity becomes continuous with that of the primary stalk cell, e.g., Dimeromyces (Thaxter 1896, 1924, 1926), Euphoriomyces Thaxt.(1931), and Trenomyces (Thaxter 1924).Basal cell m arises from the primary stalk cell, whereas cells n and n' arise from the secondary stalk cell.Thus, the lower part of cell m may more or less parallel cell VII and in species having welldefined perithecial stalks may contribute significantly to the overall length of the pedicel, as in /. peyritschii (Fig. 17, 22) and S. vulgaris (Benjamin 1968a:Fig. 51).
The stalk and basal cells of Diplomyces, Sandersoniomyces, Symplectromyces, Teratomyces, and Idiomyces persist and are well defined in age, and the perithecia of all species ofthese genera are pedicellate.Perithecial stalks of Diplomyces spp., Sandersoniomyces divaricatus, and six species of Teratomyces are relatively short compared to the total length of the perithecium.Short pedicels may prove to be characteristic of the perithecia of Diplomyces and Sandersoniomyces, but too few species of these genera are known to make one really confident of such a generalization in view of the variation in length of the primary stalk cell of three species of Teratomyces, i.e., T. mirificus (Benjamin 1968a;Thaxter 1896), T. petiolatus (Thaxter 1908), and T. insignis (Thaxter 1908).Of the known Teratomyces species, only T. petiolatus has enlarged basal and secondary stalk cells that approach those of Symplectromyces vulgaris in size compared to the perithecial body.In both of these species the combined length of these cells may equal one half or more of the length of the latter, and in the case of S. vulgaris one half or more of the length of the pedicel may consist of the combination of cells m and VII.Idiomyces peyritschii appears to be unique among known Laboulbeniales in regard to the length of basal cell m relative to the rest of the perithecium.The length of the entire perithecial body including basal cells n and n' often is less than that of cell m (Fig. 1 7), and like S. vulgaris up to one half of the length of the perithecial stalk of I. peyritschii may consist of the combination of cells m and VII (Fig. 21,22,24).
The outer walls of the perithecial body of all taxa belonging to Diplomyces, Sandersoniomyces, Symplectromyces, Teratomyces, and Idiomyces consist of four nearly vertical rows of four cells each.Except for I. peyritschii and Symplectromyces vulgaris, cells comprising the basal and subbasal tiers of wall cells in species of these genera are subequal in length and typically constitute from 80 to 85 percent of the total length of the ascigerous cavity.This proportion of the total length of the perithecial body also is characteristic of S. vulgaris, but the basal tier of wall cells in this species is about 20% longer than the subbasal.Cells of the subterminal tier range from two times to only slightly longer than those of the upper tier which always are very small and surround a strictly terminal ostiole in all cases.There are no perithecial appendages, but the upper margins of the terminal cells usually bear a small, rounded, divergent median protrusion.The trichogyne scar always lies slightly below the perithecial apex either near the base of a terminal cell or near the upper end of a subterminal cell, and it sometimes straddles the juncture between neighboring cells.Its position appears always to be the same for perithecia of a given species based on those I have studied.
The complement of ascogenic cells in the several species examined was determined by counting the number of well-developed and separated centers of ascus production (see Thaxter 1896, Pl. I, Fig. 38-39) visible in several mature perithecia.Large perithecia of Symplectromyces vulgaris contained eight ascogenous cells whereas small perithecia contained only four; perithecia of this species with six ascogenous cells were not found.Perithecia of T. mirificus, T. cf.philonthi, Diplomyces actobianus, and Sandersoniomyces divaricatus had only two ascogenous cells, a condition common to species of many genera of Laboulbeniales (Thaxter 18 96).Whether or not any phylogenetic significance can be attached to ascogenic cell number is uncertain; the number of these cells in the perithecium of a given species may be influenced simply by constraints imposed by size or even by age (Tavares 1973).
Unlike the usually elongate, slender perithecia of Teratomyces and its allies, the perithecia of/.peyritschii are comparatively short and stout.Cells of the lowermost tier of wall cells are considerably larger and often 30 to 40 percent longer than those ofthe subbasal tier (Fig. 17).However, the upper two tiers are relatively short, and the lower two tiers constitute at least three fourths of the total length of the perithecial body.The subterminal tier of cells is about twice as long as the terminal tier, and apparently the subterminal cell of the row of wall cells derived from basal cell m always subtends the trichogyne scar.Four ascogenic cells appear to be the usual number formed in perithecia of this species as earlier noted by Thaxter (1896).CONCLUSIONS 1.Previous speculations allying Symplectromyces, Teratomyces, Sandersoniomyces, and Diplomyces (Benjamin 1968a;Thaxter 1908) remain un-altered.All members of these genera have in common (1) a similar type of indeterminate receptacle consisting initially offour superposed cells in which only the upper cell, corresponding to cell III of a simple determinate receptacle, becomes further divided and gives rise to secondary appendages and perithecia, (2) similar perithecia formed in progression, and (3) closely related hosts.
2. Of these genera, Symplectromyces is perhaps the most primitive in having (1) secondary appendages lacking specialized sterile branchlets, (2) antheridia formed by direct transformation of appendage cells and developed by both primary and secondary appendages, and (3) cells of the lower tier of outer wall cells somewhat longer (ca.20%) than those of the subterminal tier.
3. Teratomyces is near Symplectromyces in the structure of its receptacle and the manner of origin of its secondary appendages and perithecia.In the formation offree phial ides by most species, Teratomyces appears to represent a well-defined and divergent line of evolution away from the other genera.However, a connection with Symplectromyces is suggested in that at least one species of Teratomyces may intergrade with that genus in its antheridial characteristics.
4. Sandersoniomyces and Diplomyces are probably the most highly evolved and the most closely allied of the four genera in view of their similar appendages and the precise manner in which the upper cells of the receptacle give rise to a reduced number of cells bearing perithecia, secondary appendages, and secondary fertile axes.Both genera have retained antheridial characteristics like those of Symplectromyces.
5. The gap between Idiomyces and the other four genera listed above probably is great enough to preclude near relationship between them.The distinctive hosts of I. peyritschii are not closely related to those of the other genera.Early divergence of these genera from a common ancestor is suggested, however, by similarities in early stages of receptacular development, especially as observed in I. peyritschii and S. vulgaris, but the mature receptacle of/.peyritschii is distinct in its type of asymmetrical development and in the origin of secondary appendages from both cells II and III of the primary receptacle.Perithecia of I. peyritschii typically arise progressively from progeny of cell II rather than cell III of the primary receptacle as in Symplectromyces and its allies; those perithecia that do develop on cells derived from cell III appear to be the result only of a teratological phenomenon.
6.Although Thaxter (1908) misinterpreted the structure of the receptacle and primary appendage of I. peyritschii, his placement of the genus in a tribe, Idiomyceteae, separate from the Teratomyceteae (for Symplectromyces, Teratomyces, and Diplomyces) is supported by my observations.
and 44, but viewed from opposite sides, are shown in Fig. lC and lD, respectively, in my 1968 study.