p 

 p 



MAISN DIATOMS OF TJTE ALBATROSS VOYAGES. 261 



pose. Kiilzing, (jreviile, and llrlglitwell, however, ignonnl (!iis latter gt-nii^, and ^o 

 pLri)etual.od tlio error. In addition to the eunlu-siun btlwtn^ii the above-nientiontnl 



( 



genera, there is a second reason for tlu^ difTieulty here encountered, and one that partly 

 explains the other; the inconstant and rather trivial generic distinction between this 

 and Coscinodiscus Ehrenb, The general structure and markings-are those of th(^ latter 

 genus, and the one striking distinguishing f<^ature, the "pseudonodule-" of Actino<y- 

 clus, is unquestionably very inconstant. Some species, like A. pyroUchnicus Deby, 

 seem fairly well marked and stable; but many others differ from welldcnown species 

 )f Coscinodiscus only in the presence (^f this i>s( udonodule, and it is not unusual to find 

 specimens both with and without this structure in the same gathering. Thus C, cur- 

 vatulvs Grun., is ji. curvaiulus .Tan., without its pseudonodule; so also C. aahtilis 

 Ehrenb. and A. subtilis Greg., C. fifsnis Norm., and A. raljsii {\\\ Smiili) Ralfs, C. 

 tubcmdatfis Grev., and^. sparsus (Greg.) Ratt.; while in II. L. Smith's type slide no. 

 12 the form he calls ''AclinorfjdffS inlcrpfrncUftns Bright.," is uniformly lacking'in the 

 pseudonodule and agrees exactly with his type slide nn. 421, marked ''rndosira 

 maAvm Grun." There might be mentioned a third rt^ason for the difTicult cliaracter 

 of this genus, namely, the lax^k of good literatun^ on the subject and especially of good 

 liguies of the species. Rattray's revision (.[ A<tinocyclusf' contains some excelU-nt 

 work, but it leaves much to b<^ <U^sired. Too many old si)ecific names have l)e(^n dis- 

 turbed; the illustrations are scanty and tno^malband the analytical key is most difli- 

 cuU to use. Schmidt's Atlas has so far practically ignored this much n(H'd(^d sub- 

 ject, an undertaking of far more importance to the science than th(^ w(\irisome 

 reillustration of species already repeatedly figured in its plates. 



Actinocyclus alienus Grun.: Van ileur. Synijp. /j^ U'^.J- V^, L^- I8S1. Ratt.Journ. 

 Quek. :\Iicr. Club II. 4: 111. 1890. WuUr, Diat. X. A. pL 85./. LU 1890. Di^Toni, 

 Svll. Alg. 2: IKJo. 1891. 



The specimens found by me agr<M» with what Gruo(»w calls variety caUJornka,^^ 



which is the type of the specif^s. 



Found at station 3301II, off tlie coast of Alaska. 



Actinocyclus crassis (W. Smithj Van Hour. Synnp. pi. JJ4. f. 6', S. 1881. De Toni, 

 fivll. AVr. 2: ll(;n. 1891. 

 Eupodisciis crassus \\\ Smith, Synop. P>rit. Diat, 1: 24. pL i-f- U - 1851?. 

 Aduioajiius suhcrassus Ratt. Journ. Quek. Micr. Club II. 4: 154. 1890. 

 Aciiriocydus circnrndaias Pant. Bacill. Ung. 1: Hf). pi. -"?./. 28. 1887. 



Van Heurck's figure is hardly typical, as the beading is radially arranged in the 

 type. Rattray's proposition to make a new species of the uonradial forms is, how- 

 ever, not to be commended. They should l)e looked upon as varieties of the above. 



Found at stati(m 3'i4fi, off the coast (tf Washington. 



Actinocyclus curvatulus Jan.; Schmidt, Atlas pi. 51. f. -?/. 1878. AVolle, Diat. Ts. A. 



pl.!).i^f.l4. 1890. 

 CoscinoflLscfis cunaliflus suhoccJJaf^'^ Grun. Donksrhr. Akad. A\ i(Mi 48'-: SI-5. pi. Jf.J .15. 



1S84. 

 AdinocyvlKS S}iho(rUatiis Ratt. Jouni. Quek. Micr. <1ui> 11. 4: 14^. 1890. 



The above differs merely iu its pseudonodule from CoscimxlLHi-ua cifrvalahi^ Grun.,'^ 

 a difiiculty that Grunow tries to av<.i<l by making ihiss])ecies a form as variety suhord- 



kit us of that species. 



Found at staffon 2807, Galapagos Islands. 



a Journ. Quek. Micr. Club 11.4: \M-2V2. pJ. //. 18iK). 

 h Van Ileur. Rynop. 125. pi 25. J. 10, 

 c Schmidt. Atlas 57./. ^J. 1878, 



