MANX DIATOMS OF THE ALBATROSS VOYAGES. 247 



which the valves turn vertically dowinvard to form a baiul, ornamented lij<e the face 

 of the valve with a rugged network; so that the two valves if placed together form 

 without the connecting girdle, a cylindrical pill box. In this they agree with Melosira' 

 though di ffering from tliat genus in mode of growth. I have adopted here the -enerally 

 achnirahle cla-ssificatlon of Rattray, idth<.ug]i in this and some other respects"it seeme 

 open to criticism, a state of things probably inevitable under any arrangement. 



Some of the members of this genus approach, througji Ihcir markings, Stictodiscus 

 and Ilemiptychus Elirenl>., and, through (heir processes, Tripodiscus Ehrenb. 

 Coscinodiscus africanus .Tan. Diat. Gaz. Exped. pi. 3./. 2, ined. Schmidt, Adas;,?. 



59. f. 24, 25. De Toni, Syll. Alg, 2: 1258. 1894. llatt. Proc. Roy. Soc. Edinb 



16: 534. 1889. 



This remarkable form shows two Pels of border markings; one of fine radial lines 

 i. e., perpendicular to the margin, the other of strong spirally and obliquely arranged 

 markings, made; very prominent in the above figure by Schmidt. I am convinced 

 those latter are an optical illusion. TJiey are caused by the fact that the actual ■ 

 border markings, the flue cross lines, do not lie parallel on the two surfaces of the 

 border, the upper and under sides, belonging t(^ the uppcT and under valves of the 

 frustule. In otlier words, the two sets of lines cross each other at a very ac ute angle. 

 This peculiarity is the result of the excentric arrangement of all the markings of the 

 valve. The two foci are not directly abov(^ each other; and therefore the lines radi- 

 ating from them, including the fine lines on the border, are not parallel. As a con- 

 se<pience the false spiral markings app(>ar Ix-st under low powers of magnification 

 or when slightly out of focus. Under a oiu-twelfth inch objective and with careful 

 focusing they are not to be seen. 



The curious excentric arrangt'ment of the network and the spiral appearance of 

 the bord(-r seem like an abiiormalily. Rut (he species is very true to its type. My 

 specimens come from the; same locality as the original, the Galapagos Islands. They 

 are another illustration of the value of the diatoms for indicating locality, fastra- 

 cane's "var. rotunda'' of this species," though recognized by De Toni, has nothing 



to recommend such an assignment except the mere accident of an excentric arrange- 

 ment of markings. 



Found at station 2807, Galapagos Islands. 



Coscinodiscus asteromphalus Ehrcnl). Rer. Akad. Wiss. Bcrl. 1844: 77 1S45- 

 Mikrog. pi. 18. f. 45, pi. 33, XV.f. 7. 1854. Priteli. ITi?t. Infus. ed. 4. 828'. ISGl' 

 Grun. Denkschr. Akad. 'A'ien 48': 78. pi. 3.f. 9. 1884. Pant. Beltr. Bacill Ung 

 1: 73. pi 17. f. 153. 188G. Ratt. Proc. Roy. Soc. Edinb. 16: 549. 1889. Schmidt 

 Atlas pi. 63. 12. 1878; pi. 113. f. 22-23. 1888. Cleve Si Moll, type no. 57. 164.' 

 Jan. Diat. Gaz. Exped. pi. 4- J. 9. Van Heur. Syiiop. pi. 128. f. 1-3, 5- pi 130 

 f. 1-2, 5-6. 1881. De Toni, Syll. Alg. 2: 12G8. 1894. 



Cosmwdiscas omphnlanlhiis Ehrenb. Ber. Akad. Wiss. Berl. 1844: 206. 1845 

 Pritcli. Hist. Infus. ed. 4. 828. 18G1. Cleve & Moll, type no. 57, 215. H. L. "Smithj 

 Sp. Diat. Typ. no. 95 (not Schmidt, Atlas pi. 63. f. 2. 1878). 



CosrinodiscHS oculiis-iridis Ehrenb.; Schmidt, Atlas pi 63. f. 5. 1878. 



Coscinodiscus centralis Ehrenb.; Schmidt, Atlas pZ. 63./. 1. 1878., 



Rattray says, "This species is distinguished by the (>vident puncta on the mark- 

 ings." But many other species share this character, as C. marginatus Ehrenb. and " 

 C. rohuaius Grev. Rails .says, 6 ''It differs from \\\<- othcT species with stellate umbili^ 

 cus by its minutely punctate cellules." These two distinctive characters should be 

 taken togf'ther. 



Foun<l at stations 2929, 3570, 228711, 303511, 4023II, 4025II, off Calif(^rnia to Bering 

 Sea. 



aCastr. Rep. Voy. Chall. Rot. 2: 159. pi 24. f. 3. 188G 

 *Pritch. Hist. Infus. ed. 4. 828. 1861. 



31713 — VOL 10, FT 5—07 3 



