316 COXTRTBUTTONS FROM THE NATIONAL HERBARIUM, 



for tlio sopanitiim l>y Sclnnidt of Ids P. injlexa. There is l>ul little dcMilit (tf P. robasta 

 belonging Ikto; the slight difforc^nce in heading iirnounts to nothing, as this s])('C'ies 

 is most Yarial)le in tluit repp<Htj and the lighter and more curved septa appear to me 

 uniini)ortant. I have, however, followed this .^yn-Jiiym with a qucFtion mark. 

 Found at station 2807, Gala])agos I:jlaiids. 



HEMIDISCUS WaU. 



Hemidiscus Wall. Trans. Micr. Sor, Lomh n. s. 8: 42. pL '^.f. /M- 18G0. Priteh. Illst. 



Infus. ed. 852. 180)1. 

 Eiiodia T^ail.; Pritch. Hist. Tnfus. ed. 4. 852. ISfil. Pvahh. Fl. Fur. Alg. 1: 317. 



18G4. Castr. Rep, Voy. Thall. Bot. 2: 148, 188(). Van TIeur. Treat. Diat. 538. 



1890. Pe Toni, Syll. Alg. 2: 1324. 1894, 11. L. Smitli, The Lens 1: 92. 1872. 

 Dichomeiis Ehrenh, Phys. Abh. Akad. Wiss. lierl. 1861: 294. 18G2; 1872: 205. 



pi 9 J. S. 1873. 

 Falmeria Gnn'.; Grev. Ann. Mag. Xat. IlisL IIL 16: pi. 5. f. I-4. 18G5. Van 



Heur. Treat. Diat. 538./ 'j^SO. 1S9G. IL L. Smith, Sp. Diat, Typ. no. 387, 1874. 

 Goniofhecivm Ehrenh. ^Lkrng. pi. 33. XVIIl. /. 4- 1854, in part. Griff. & Ileiif. 



Mier. Diet. ed. 3. 347. pi 4^-J- 18. 1875, 



Most diatomists have retainc^d the name iMuxlia Bailey, ]K)twit]istanding the fact 

 that Hemidiscus of "Wallieh is a year oLLt. Besides this ol)jeetion to the name tlu^re 

 is a still stronger one in th(^ fact that the name Euodia was us(m1 by Forster in 1770 

 for ageniisof flowering })lants of the family Rutaeeae. That Euodia Bailey and ilemi- 

 diseus WalL are synonymous is certain, the only mark of distinction being the p?eu- 

 donodule observed near the ventral margin of thi,' \al ve by "Wallich. As Ralfs remarks,'^ 

 this may have been overlooked by Bailey; ur, as is more likely, it was absent from his 

 specinu^n. It is not infrequ(^nt to fmd the same forms with or without the ]tseudo- 

 nodulcj or to fmd a complete frnstule with on(^ valve Ijearing a pseudonodule and the 

 otlicM- lacking it. It is also impossilde to hohl Greville's Palm<Tiaas a separate genus. 

 The indefinite hyaline central area is the only real difference it possesses, a charac- 

 teristic too trivial to warrant its separation, as on the same ba^is we should have to 

 divide many other genera, such as Coscinodiscus. The othc^r quality claimed for 

 Falmeria, namelyj fine submarginal spines, with lines from these i-Tinning i-adially to 

 the center, is not at all peculiar to Greville's ;s|)ecimen. AVallich lefers to the mar- 

 ginal processes as "marginal puncta'^ and figures them on the dorsal side. A careful 

 examinatii>n of Greville's Pahneria in the excellent specimens of IL L. Smith's 

 type slides no. 387 convinces me it is withoul th(^ distincti(uis necessary for separate 

 generic rank; thougli it sliould be classed as a sei)arate specic^s, viz, Hemidiscus laud 

 maniaiuis (Grev.) Mann.^ On the other hand, it should Ito Ikto noted that the strik- 

 ing build of the members of this genus leads to the too easy ('(Uiclusion that all 

 cuncuform frustules bearing a reasonably close reseml)lance to the original species, 

 Hemidiscus ciinciformis Wall., must Ite mere varieties of it. Of course such a conclu- 

 sion would l>e no more justified than to look upon all circular diatoms covered with a 

 delicate, radially arranged network of hexagons as varieties of Coscinodiscus radiatns 

 Ehrenb. I have, therefore, below given to two specimens found by me the sei)arate 

 specific names bestowed by Castracane, although their gent^ial similarity to Wallich's 

 original form is appart^nt, 



Hemidiscus cuneiforniis Wall. Trans. Micr. Soc. Lond, n. s. 8: 42. pi 2. f, 3-4 

 ISGO. Moeb. Diat.-taf. pi Jl.f. 3-4- 1890. Pritch. Hist. Imus. ed. 4. 853. pL 6, 



r 14. i8r>]. 



"Pritch. Hist. Infus. ed. 4. 853. 18G1. 



^Euodia hardmaniana (Grev.) H. L. Smith; Hab. Cat. 132. 1877 



