Fraxinus Parryi, Nom. Nov., of NW Baja California, Mexico

The common ash of NW Baja California is usually called Fraxinus trifoliata, as in Wiggins' (1980) flora of Baja California. That name is a misspelling of F. trifoliolata, which is a later homonym of F. trifoliolata W. W. Smith (1916). The Baja Californian ash is therefore renamed Fraxinus parryi. It has also been misidentified as F. jonesii. The common name is "crucecilla". This ash barely extends into San Diego County, California.


INTRODUCTION
Two species of Fraxinus grow in northwest Baja California, Mexico: (I) two forms of F. velutina Torr., which are rather uncommon along streams southward to below the middle of the peninsula, and (2) F. dipetala var. trifoliolata Torr. or F. "trifoliata" (Torr.) Lewis and Epling, which is widespread and common in the northwest. It is the latter species that I am concerned with here.

RESULTS AND DISCUSSION
Fraxinus trifoliolata Torrey (1859) named F. dipetala H. & A. var.? trifoliolata (sic) from a specimen collected by C. C. Parry in 1850 in "sterile mountains a few miles south of the Mexican line in Lower California". Torrey remarked that since the flowers of this ash and the fruit of F. dipetala were unknown, he was uncertain whether the two were distinct or were only extreme forms of one species. Sudworth (1908) and Standley (1924) listed the Baja Californian shrub as F. dipetala trifoliolata, Little (1953) called it variety trifoliolata, and Murray (1985) made it subspecies trifoliolata. Miller (1955) and Little (1979) called it a synonym of F. dipetala. Goldman (1916) misidentified it as F. attenuata M. E. Jones. Lewis and Epling (1940) described this Baja Californian ash and compared it in detail with F. dipetala H . & A. of Alta California, calling it a separate species rather than a variety. They found the leaves conspicuously different, with 1-3 leaflets rather than (3-) 5-9 and so averaging only 4.5-cm long as compared to 11 .2 ern, the terminal leaflet averaging 1.5 times as long as the laterals rather than just as long, and the margins subentire rather than serrate. They also found small floral differences, notably the petals averaging 3.3 mm wide, rather th an 2.2 mm; the pet-als and filaments fused into a tube 0.6-1.5 mm long, rather than fused lightly at the base or rarely forming a tube ; the fi laments 0.4-1.0 mm long above the fusion , rather than 0.8-2.8 mm long; the style averaging as long as the stigma, rather than 1.4 times as long. Wiggins (1964Wiggins ( , 1980 and others likewise have treated this ash as a distinct species, and I agree. Wiggins (1980) showed a drawing of it as his figure 413.
Lewis and Epling (and likewise Wiggins and others) wrote the name as F. "trifoliata '' (Torr.) Lewis & Epl. Because they were making a new combination, however, and not intentionally coin ing a new name, their name is correctly spelled" F. trifoliolata"; as shown, for example, by Little (1953Little ( , 1979. It is therefore a later homonym of F. trifoliolata W. W. Smith (1916), a Chinese ash. Since this Baja Californian species thus seems to have no legitimate name, I propose to call it Fraxinus parryi, nom . nov., based on F. dipetala var.? trifoliolata Torr. (1859: 167); holotype: Lower California , lat. 32°, C. C . Parry, 1850, Torrey Herbarium (NY). It is named for the first collector, C. C. Parry (1823-1890), botanist and plant explorer.

Fraxinus parryi
Fraxinus parryi is a shrub or small tree widespread in NW Baja California (Fig. I), from near sea level to at least 950 m near Valle Trinidad toward the west base of the Sierra Juarez, and to 1200 m near EI Socorro, in the western foothills of the Sierra San Pedro Martir: and from just north of the Mexican boundary near Lyons Valley, San Diego County, California (Beauchamp 1986 . . above most or all the other shrubs . Lewis and Epling (1940) wrote that it does not usually grow within typical chaparral but is generally peripheral or associated with, but apparently not genuinely a part of, the coastal sage. Since he said flowers and fruit were unknown, both spec ime ns must be sterile. Hence it may seem reckless for him to have based a new species on these collections or even to call them the same species. Although the epithet jonesii might suggest that the Jones collec-tion be selected as lectotype, that is not an automatic choice, Marcus E. Jones's own herbarium at Pomona has one specimen (POM 76344) of Jones 3740, Valley of Palms, Lower California, April 8, 1882. This has glabrous leaves with 1-3 unequal leaflets, the leafletsespecially the terminal one-markedly attenuate-petiolulate: and it bears flowers with two petals: clearly it is F. parryi. However, Lingelsheim's description of F. jonesii includes the phrases "folia 1-2-juga" and "foliola ... sessilia vel fere sessilia utrinque, imprimis subtus, long ius albido-pilosa" : His rather crude figure 9G shows a leaf with 5 subequal and subsessile leaflets and with conspicuous spreading short pubescence. Lingelsheim's description and figure could not be based on Jones 3740 as that is represented at POM and also at CAS and OS; so F. jonesii Lingels. clearly has nothing to do with F. parryi.
On the other hand, remarkably, Lingelsheim (1920: 33) based F. schiedeana var. palmarum Lingelsh. on Jones 3740 ex parte from the Valley of Palms, and his figure 9A of this variety shows a flowering plant like F. parryi, though one leaf has five leaflets! Again he cited no herbarium. Lingelsheim (1920: 32) also listed F. dipetala var. trifoliolata Torr., which he said was unknown to him but was perhaps an independent species.
What could be the source of the confusion? At or near the Valley of Palms (Valle de las Palmas) on April 8, 1882, Marcus E. Jones collected not only F. parryi (number 3740) but also two forms of F. velutina. His number 3741 he labeled as F. pistaciaefolia Torr. var. but later (Jones 1908) named F. attenuata n. sp . Specimens I have seen of 3741 (CAS, OS, POM) have 3-5 attenuate glabrous leaflets. Jones also made two unnumbered collections of a pubescent form of F. velutina with 5-7 wider leaflets, one (OS, POM) also on April 8 but labeled "Northern Lower California", the other (CAS) at Valley of Palms but on the return trip, April 15. Possibly the syntype of F. jonesii that Lingelsheim listed as Jones 3740 ex parte was mislabeled and instead corresponded to the pubescent second unnumbered collection of F. velutina-though the pubescence is not spreading as in Lingelsheim's figure 9G, and the leaflets number more than 3-5 . The location of this specimen is unknown.
Lingelsheim's other syntype of Fraxinus jonesii was Pringle 137 ex parte, whose location also is unknown. According to Lingelsheim (1920: 26), Pringle 137 otherwise is F. cuspidata Torr. from Chihuahua, which seems to differ from F. jonesii in its glabrous longpetiolulate and long-acuminate leaflets. But what Pringle 137 ex parte may be remains a mystery. CONCLUSIONS Thus known specimens of Jones 3740 and Pringle 137 are both quite different from the description and drawing of F. jonesii. By a remarkable and unfortunate coincidence, both syntypes of F. jonesii, cited as Jones 3740 ex parte and Pringle 137 ex parte, evidently were labeled incorrectly as to the collector's field number, making those collections even more difficult to track down. Possibly these two syntypes of F. jonesii were lost with other specimens, as in the Berlin herbarium during World War II. On the other hand, perhaps as more collections are entered into databases, these two lost syntypes may come to light, the name of Fraxinus jonesii can be lectotypified, and this name can find its proper place in synonymy. What is important here, however, is that F. jonesii clearly has nothing to do with F. parryi and so can be excluded from the discussion of this Baja Californian ash.