^ 



MANN — JJIATOMS OF THE ALBATROSS VOYAGES. 279 



fartluT apart tluin elsowhorf^ giving to tbr intrrsj)ac(' [he appoaranco of a bright, nar- 

 row ring; pmrosses two or 1]irc(^, .^tout, i^pherical, pla('(Hl at about oiK'-wixth a radiurf 

 from tlu' margin; th(^ l)ast? of each .surrouiub'd by a .^mall hyaline area, from whii'li an 

 obs(iir(^ hyabnc lino extends to the center; ))order very narrow, minutely l)eade(b 

 Diameter of valve O.OGr>-to 0.133 mm. 



Type in the U, S. Xatiemd Museum, No. 51)0131 , from sUition 4505II, Santa Ouz 

 LighMiouise, MontiTey Bay, Cab, 1904; lOlalhems. 



Tlu^ efm<'entric beading of lliis sp(H'ies reminds on(^ of .1. brownii Norm., though it 

 i.s much mr^re pronounced her<\ The species differs frfmi ^1. brownii in the convexity 

 of the valvo, the character of llu> processes, the liorderj etc. The latter ii> al^o a mucli 

 more di'licate species than Ibis massive one. 



Although Rattray has used the aliovc sj)eeili(: name for an Actinocyclus, I have 

 applied it here because it is so aptly descriptive of this diatom's most striking 

 characteristic. 



Tripodiscus cosmiodiscus ^hinn, sp. nov. Platk I/IV, yiorRK 4. 



Valve circular^ nearly flat, beaded; lieading liiinute, closely and evenly set in 

 radiating lines^ except where these lines are slightly bowed around the twelve some- 

 what protuberant portions of the valve bearing ihe twelve processes; no hyaline cen- 

 tral area; margin obscure; each of the twelve ])ro( esses consisting of a j)rolongation of 

 a single row of beads, this within one-sixth a radius of the margin becoming a narn>w 

 hyaline ridge crowning the twelve slightly protul^erant portions of the valve and ter- 

 minating in an exceedingly small but evidently spherical process. Aspect of the 

 valve somewhat like Greville's Cosmiodiscus hu/hadmsisJ^ 



Diameter of valve, 0.0!>5 mm.; beads 130 to 135 in (l.i mm. 



Type in the U. S. National Museum, No. 590l:^\ from station 4029II, Bering Sea, June 

 27, 1900; 913 fathoms, bottom of gray sand and clay. 



Tripodiscus kinkeri (Schmidt) Mann. 



Aidacodiscus kinhrn Schmidt^ Atlas ;^/. 106./. -j ;>. iSSO. 



Anldcodiscus margaritace}!^ hinkcri Ratt. Journ. Roy. Micr. Soc. 8^: 352. 1888. 



Tlie dominant form of the present genus is probably to be t^oon in A. manjaritaceus 

 Ralfs; and under that speeilic name Jiattray has grouped a host of forms considered 

 as se]>arate species by other authors. ?j This condensation was much needed, and has 

 been well done. Ibit in a fi^w instances it seems to lia\e l>ecn carried too far. The 

 above case is an example. S(hmidt s species has a strong general resend>lance to 

 both A. margaritacius Ralfs and A, crux Ehrenb.; but I think to make it a vari(^ty 

 is hardly justified. Schmidt'sform is a far flatter specimen than Ralfs's species. As 

 the <lifference between varieties and species is ])ractically a thing of personal opinion, 

 the recognition of Schmidt's species is simply an expression of the fact that I look upon 

 it as a justified expedient for distinguishing between this form and others that are 

 called .1. vKirgaritaccus Ralfs. It may not be cnit of place here to add that the multi- 

 plication of named varieties is, I thhik, to be avoided wherever possible. As a rule 

 if a new form is so different from an already existing sj)ecies as to retpnre taxononn'c 

 recognition, it is safe to louk u[)oii it as a new si>ecies. The gradations of tlie diatoms 

 are almost limitless, and named varieties can be multiplied ad li!)itum. Tn this 

 w"ork I have omitted these whenever possible and liave added none to the list myself. 

 The remarks on this subject by llev. William Smillit- are worthy of consideration. 



Found at station 3(19G, off Honshu Island^ Japan. 



aTrans. Micr. Soc. Loud. n. s. 14: 79. pi ,V./. i;?. 18GG 



2> Journ. Roy. Micr. Soc. 8': 351. 1888. 



c Quart. Journ. :Micr. Sci. 3: 130^135. L855. 



31713 — VOL 10, FT Ti— 07 r> 



