250 CO^sTRTBUTIONS TROM THE NATIONAL IIEHBAKTUM. 



This vory delicate spccios is with some dinicully .soparatcnl from C. centralis Ehrenb. 

 W. Smitli claims that its markingt^ arc so mucli finer that Ehrenberg could not have 

 seen them with his mierosrope. This is indeed the difference. But it is a question 

 if this offsots their close rt^semblance, in both having a central rosette of very large 

 cells surrounded by unusually fine markings, in both having the two peculiar mar- 

 ginal processes which led Oreville to class them as Eupodiscus, etc. The separation 

 is at least open to question. Grevilb^'s l>eautiful figure of E. jonesianus is inaccurate, 

 and llopcr's figure emi)hasizes'too strongly the radiating lim-s. 



Found at stations 2807, 451GII, Galapagos Islands, and off Lower California. 



/■ 



/.6M888. 



/■ 



Gaz/Exped- pi. 2./. 7, p?. 5./ l^ 3,S, pi 6 J. 2, pl.20.f. 17. Tleve & Moll, 

 types no, 57, 154, 102, 1G4, 270,319. Ilatt. Proc. lloy.Soc. Edinl>. 16: 180. 1889. 

 DeToni, Sylh Alg. 2: 122G. 1894. 

 Cosdnodm-us {Odonioducua) curvatulus Grun.; Cleve & Grun. Sv. Vet. Akad. Hand!. 



/ U9, 1880. rh-ve in Xordensk. Vega Exped. 3: 488. 1883. 

 iaqhii Pant. Peitr. Pacill. ling. 1: 74. pL 15, f. 133, pi, 28. f. 



188G. 



In my specimens the curvature of the fascicles is so slight as to be easily overlooked, 

 a characteristic whidi, though emphasi/.ed in the name, is unimportant and variable. 

 There can l>c no doubt that this form and Actinocychfs cnrvatidus Jan." ar<^ the same 

 diatom except for the accident of a pseudonodulc in the latter, I do not regard the 

 figures by Janisch ^ as behaigitig ht^n*, though Rattray so classifies them. Ilisroference 

 to II. L* Smith type no, 99 as containing this species under the name of Odonlodiscus 

 carvatuliiH Grun. is wrong, so far as the slide I have is concerned. It comes from 

 Japan and is marked C. sdnUllans Grev., and although it shows abundant examples of 

 C. subiiliH Ehrenb., whi(4i has some slight reseml)lanc(^ to this species, it is a 

 quite distinct form, a,s Rattray points out. It is proboble that Rattray's slide of 

 Smith's no. 99 shows this species; ))ut the reference is faulty as a general reference. 

 Pantocsek\s figures of C. szontayhii are so utterly different from this species that they 

 an^ wors(^ than useless. ]?y reading his description it is, hctwever, plain that Rattray 

 is correct in uniting this form with the above. It would be well for diatomists to 

 either ft>rogo the luxury of illustrations or make them near enough like the objects in 

 nature to be capable of recognition. 



Found at stations :^G04, -I022H, Bering Sea. 



Coscinodiscus decrescens Grun.; Schmidt, Atlas pL 6L f. 8-10, 15. 1878. Grun. 

 Denkschr. Akad. Wien 48': 80. pi S. C, f. 11, IS, 1884, Ratt. Proc. Roy. Soc. 

 Edinl). 16: 525. 1889. Do Toni, Syll. Alg. 2: 1252. 1894, 



Rattray, followed l>y De Toni, includes in the above C. heleroponis Ehrenb.'* and 

 C. argus Grun.'^ I think tlu^ assignnu^nt in Schmidt is miicli better. Rattray places 

 a question nuirk aftor Schmidt's plale Gl, figure 10, whicli, if an example of this species, 

 nuist certainly be looked on as abnormal. Castracane has assigned this name to a 

 totally different diatt)m,e for wdiicli Rattray proposes the name C. rninucns. 



Found at station 33G1II, Bering Sea. 



Coscinodiscus deformatus Mann, sp. nov. Pi-ate XLVIII, figures 1, 2. 



Valves nearly fiat for four-fifths of iho radius, thence curving slightly downward; 



network radially arranged, somewhat irregular at the center, but with no umbilicus; 



I _ . _ -"__T " — "" • _ ■_ ■■_^- ~ • ~l • ^i~ —^LJ — T. . m ■■ L ■ W ■ |__|_^ ^^ l__^J- 



«Cf. Schmidt's figures above with his pi 57, f, 31, 



i^ Jan. Dial. Gaz. Expod. pi l.f. 6\ pi 4,f^ 4- 



c Schmidt, Atlas pi 61,/, 6-7, 1878, 



fiSchmi(!t, Atlas pL 113,/. 7, 1888. 



^Ca^tr. Rep. Voy. Chall. Rot. 2: 159. pi 12./. U^ 1886. 



