310 CONTRTBUTTONS l^ROM THE KATIONAL HERBARIUM. 



Tiiccralmm tridad ijliim l^riglit.'^ has long-attenuated processoR, (l(^lirate reticulation, 

 and a strikingly distinct border. Hall's, Dc Toni, and Boyer unite it with the above, 

 but Schmidt and others consider theui di.stinct. I agree with this latter view. I also 

 exclude T. pileus EluL'nh.,^' which Schmidt ^ looks upon as synonymous with T, spi- 

 nosum Hail. The resemblance is not worth considering. Grunow's claim that 

 T. spinosum Bail, is only a triangular form of Biddiilphia yramdala Roper is equally 

 untenable. The borders arc very different, and t.h<^ whole build of B. grmiulata, 

 especially its reticiilation, is much finer and more delicate than in 7\ spinosum. 



As th(^ name Biddulphin spinosa has been applit^d by Grevilh^^^ 1o a quite different 

 diatom and as I consider Grunow's assignment of this species of Greville's to Denti- 

 cella ^ to be made on iuadc(iuate grounds, the original specific name of Cailcy is pre- 

 empted, and llie choice lies between Bailey's T. setigenini and Roper's T. annaliini. 

 Both were puldished in 1S54; but as Baih^y's article appeared in February and R4:)per'8 

 somewhere near th(; close of tin? year, I have selected the nanu^ assigned by Bail(*y. 

 My form is a large and elogant variety of this varia1)le species. It shows a distinct 

 inner triangular area syminetric^rl with the onlcr triangloj its reticulation showing no 

 radiation; but outside of tliis triangle the reticulation i? radial, running vertically to 

 the edge, while a few rows of the network fonn bi'oad lines running from each apex of 

 the inner triangle to the base of each horn-like j)r(jcess in the angles of the outer tri- 

 angle. No sut\u*e exists between these vaiious jxirtions of the valve, but the pattern 

 is mad(^ very (evident by tlu^ arrangcuncnt of tli(^ reticidation. 



Found at station 2807, Galapagos Islands. 



Biddulphia sliadboltiana (Grev.) ^[ann. 



Tnceratmm ? gibbosutn Harv. & Bail. Proe. Acad. Phila. 6; ISl. pi. 9.f. 32. 1858; 



7: 430. 1854. Schmidt, Atlas pi. 80. f. 1S-15, 17, t>l. 1682. 

 Triceratium orbindatum Shadb. err. det. Bright. Quart. Journ. Micr. Sci.4: 270. pL 



17. f, 20, 185G. 

 Triceraiiiitn sluidboltiarmm Grev. Trans. Micr. Soc. Loud. n. s. 10: 28. 18G2. Van 



Ileur. Synop. pi 108. f. 5-7. 1881. Schmidt, Atlas pi 80. J. 18-20. 1882. Do 



Toni, SylL Alg, 2: 954. 1894. 

 Tncerallum elonyatum Grun. in Schmidt, Atlas pi SO.f. 12. 1882. 

 Lamprisciis kittoni Schmidt, Atlas pi 80. f, 11, 1S82. 



I think there is sufficient ground for rejecting Triceratium orbiculatuni Shadb. in 

 connection with this species. Greville explains with care / that I3jightwell confused 

 another form with the original J\ orbiculatiivi of Shadbolt. This is borne out by 

 Shadbolt's description and Ilgure,f/ the ligure lacing ri^produced by^^oebius,^ Shad- 

 bolt mentions no spines, and they are omitted from the figure by Tuffin West, wliich 

 clearly argu(\s they were not there. It is true, as Beyer i has pointed out, that Jiright- 

 well's species may be with or without spines, "a statement tliat I can confirm. But 

 though Brightwell's species may n^scmble Shadl)olt's in not having spines, that does 

 not mean that Shadbolt's species resembles Brightwell's further than in this negative 



quality. Greville, in conferring the name ahadboUianuni on the misnamed species 

 of Brightwell, clearly emphasizes their differences and Ralfs J repeats this distinction. 



^ Quart. Journ. Micr. Sci. 1: 248. pL 4^f. S. 1853. 



h Ehrenb. Mila'og. pi 19, f. 18. 1854. 



c Schmidt, Atlas pi 87. f. 18. 1885. 



d Trans. Micr. Soc. Lend, n, s. 13: 6. pi l.f, 3, 1865. 



e Denkschr, Akad. Wien 48-: 58. 1884. 



/Trans. Micr. Soc. Lond. n. s, 10: 28. 1862. 



^ Trans. Micr. Soc. Lond. n. s. 2: 14. pi l.f. 6. 1854. 



^Moeb. Diet.-taf. pi 3./. 6. 1890. 



iProc. Acad. Phila. 1900: 710. 1901. 



;Pritch. Hist. Infus. ed. 4. 853. 1861. 



