MANN— DIATOMS OF THE V.LBATROSS VOYAGES. 315 



arc synonymous with //. triquetra: II. compressa Wall.,a //. mauritiana Berger,6 //. bar- 

 ijiuiii Pant. c Triceratium javanicum Cleve. d Bui there is doubt whether these belong 

 here They would be \ ery widely dr\ ergent varieties, and it seems be i to leave them 

 in a combination by themselves under the specific name triqm tra. '/'. Pli wrodesmium i 

 brebissonii km/... though an evidenl Terpsinoe, is also in strong enough contrast to 

 Btand specifically. I have based this conclusion on a careful comparison of the two 

 species in 11. I.. Smith's type no. 388 /'. brebissonii and no. 590 / mu ica and on 

 the agreement of these with the figures of Ralfs in Pritchard'e History of [nfusoria.< 

 Found at Btation 2807, Galapagos [slands. 



PORPEIA Bail. 



PorpeiaBaiL; Pritch. Hist. Infus. ed. L. 850. pi. 6.f. 6. 1861. Rabh. I'l. Eur. Ug. 1: 



315. L864. Grev. Trans. Micr. [Soc. Lond. n. s. 13: 52. pi. 6. /. 18 11. L865 De 



Tom. Sn 11. Alg. 2: 893. L892. Van Heur. Treat. Diat. 177../'. \08. L896. Boyer, 



Proc. Acad. Phila. 1900: 731. L901. 

 This genus is a connecting link between Biddulphia< rray and Hemiaulus Ehrenb. 

 lis difference from the former is largely a matter of greater massiveness of frusl ale and 

 the internal sepia: its <li fference from the latter is in the absence of st raighl or curved 

 spines at the ends of its processes and in its septa. In fad . a Btudy of the zonal view 

 of these genera will shew hew Little is needed to turn one into the other. Schmidt in his 

 Atlas i has figured a species which, as hesays, is highly interesting, a Eemiaulus-Porpeia 

 with a rudimentary curved spine at the end of each process. Or, if we take figure l_ 

 of the same plate and broaden laterally its central elevation till ii presses on the two 

 processes, we have a Porpeia, and the two sutures between the central area and the 

 processes reproduce exactly the septa of Porpeia. But though this genus could in a 

 broad sense be muted with either of the ether two, Eemiaulus could nol be included 

 in Biddulphia wit hen t abandoning en1 irely our presenl conception of it . For classi- 

 fication it is. therefore, certainly besl to retain this genus, as is universally dene. 

 Porpeia quadriceps B ail.: Pritch. Hist. [nfus. ed. 4. 850. pi. 6. f. 6. L861. Van Eeur. 



Synop. pi. 95bis. f. 12 I ',. L881. Grev. Trans. Micr. Soc. Lend. n. s. 13: 52. pi. 



6.f. is 19(pl.8. J. IS doubtful). 1865. Schmidt. Atlas pi. 142. f. 38, tf-56. L889. 



Van Heur. Treat. Diat. 477. 1896. DeToni, Syll. Alg. 2: 893. I MM. Moeb. Diat.- 



taf. pi. 67. /. 18-19. (pi. 68. J. IS doubtful). L881. Well,.. Dial. X. A. pi. 66. /. 



25 (/. 30 doubtful). 1890. Boyer, Proc. Acad. Phila. 1900: 7:11. L901. 

 Porpeia quail rata Grev. Tram. Micr. Sec. Lond. n. s. 13: 53. pi. 6. f. 20. L865. 



Moeb. Diat.-taf. pi. 67. f. 20. L890. DeToni, Syll. Alg. 2: 893. 1894. Van Eeur. 



Synop. j>l. Uobis.f. 15. L881; Treat. Diat. 177../'. 208. L896. Wolle, Diat. X. A. pi. 



66./.33.1S90. Boyer, Proc. Acad. Phila. 1900: 732. L901. 

 Porpeia ornata Grev. Trans. Micr. Sec. Lend. n. s. 13: 53. pi. 6. f. 21. L865. Moeb. 



Diat.-taf. pi 67. f. /I. L890. DeToni, Syll. Alg. 2: 894. L894. 

 Porpeia robusta Truan & Win. Dial. Eayti L8. pi S. f. 23. L880? De Toni, 



Syll. Alg. 2: 894. L894. 

 Porpeia inflexa Schmidt. Atlas pi. r,!./. :,s. L889. 



1 think Schmidt ih<i'< rightly in combining Id- fourteen figures above cited under 

 this single species. They practically are identical; figure 17 is what Greville calls /'. 

 quadrata, and his figures 54and 56 are Greville's /'. ornata. But 1 can see no reason 



"Op. en. 252. pi /;../'. ; 12. 



6 Le Diatomiste 1: 31. />l. 51. pi 5.f. 8. L890. 



c Pant. Beitr. Bacill. [Jng. 2: 82. pi L889. 



dBih. Sv. Vet. Akad. Ilandl. 18"': 24. pi 6.f. 75. L881. 



< Ed. l. 859, pi. n. J. 47. 1861. 



/PL 142/. 67. 



