NEW FORMS OF MARINE DIATOMACE^. 493 



central nodule indefinite. Terminal nodules considerably within the margin, 

 small. The two parts of the median line terminate in the middle in small rounded 

 expansions, but do not meet. Stria? delicate, but sharp, transverse in the middle, 

 and gradually more and more curved towards the apices, where they become nearly 

 vertical. They are not so close together as in some forms in which they are 

 equally delicate, and there are about 54 in 0-001". Valve thin, flat, hyaline. 



This remarkable form occurs in Professor Allman's Lamlash Bay dredging, 

 where, however, it is rather scarce ; also in that of Mr Miles. 



29. Cocconeis splendida, n. sp. PL IX., fig. 29. Form, a pure broad oval. 

 Length about 0*0044"; breadth about 00039". Valve strong, and very richly 

 marked with strise, which are highly inclined and curved near the apices. These 

 striae are coarse, about 14 in 0001", and are formed of granules, which gradually 

 diminish in size towards the median line. The four or five outer granules of each 

 of the strise are set as closely together as possible, while the rest are separate. This 

 gives the appearance of a broad marginal band. There is a small, nearly square 

 blank at the centre, which is no doubt the indefinite central nodule. The two 

 halves of the median line are strong, somewhat bent at the terminal ends, where 

 they form elongated expansions, lying just within the dense marginal band. The 

 central ends terminate in small expansions, which lie at the upper and under 

 edges of the central blank. 



This beautiful form occurs in Lamlash Bay, but it is hitherto scarce. No doubt 

 it will some day be found more abundantly. It is, with the two preceding forms, 

 remarkable for the size it attains, being the largest Cocconeis yet described, while 

 C. major and C. pseudomarginata are but little below it in this respect, and even 

 C. nitida is unusually large for this genus. 



GROUP III. 



Filamentous Forms. 



Of this class of forms, the number is considerable. It is worthy of remark, 

 that most of them belong, so far as I am able to judge, to the genus Denticula, 

 which hitherto has yielded only fresh- water species. But there are six of which 

 the genus is doubtful ; partly because the F.V. alone is as yet known, which is the 

 case in four of them ; partly because, if not Denticulse, they cannot well be re- 

 ferred to any of the genera in Smith's Synopsis. One remarkable species, the 

 genus of which is still uncertain, I have been compelled to remove from Himan- 

 tidium, in which genus Professor Smith had provisionally arranged it. The forms 

 of this group are 14 ; viz. : — 



30. Denticula (?) interrupta, n. sp. 



31. ... (?) capitata, n. sp. 



32. ... (?) ornata, n. sp. 



33. ... (?) lsevis, n. sp. 



34. Denticula nana, n. sp. 



35. . . . minor, n. sp. 



36. ... distans, n. sp. 



37. ... staurophora, n. sp. 



VOL. XXL PART IV. 6 R 



