OBSERVATIONS ON PLANT PLANKTON. 393 



the bodies of small Crustacea ; in this latter case, they would have 

 been broken up into minute fragments.* 



During summer, I made a long series of attempts to discover 

 whether the species of CeratUun, then so abundant in the sea, 

 furnished a constituent of the food of small Crustacea, but without 

 result. In the first place, Cerathim could hardly be traced so easily 

 as diatoms, and the matter is one of difficulty. Direct observations 

 also yielded no result. I made an estimate of the relative abun- 

 dance of diatoms, of species of Ceratium, and of animal life in all 

 the captures, and I found a certain constancy in the proportions in 

 the great majority of cases. When there were plenty of diatoms, 

 there were alivays plenty of Crustacea; when Ceratium predominated, 

 and diatoms were scarce, there was little animal life. (In Loch 

 Fyne, in July, I made two almost pure captures of Ceratium Tripos 

 — merely one or two diatoms and copepoda mixed with it.) These 

 facts, however, are merely suggestive. There is one more suggestion 

 I should like to make. The spines of Ceratium may be a protection 

 against being eaten. I could not fail to remark that there were no 

 traces of spiny diatoms, such as the species of Clmtoceros, having 

 been eaten by small Crustacea, though they were abundant in the 

 sea. Against these negative observations, however, may be placed 

 the fact that I found Chcetoeeros in the young sand-eels. Though 

 one can suggest other uses for the spines and sharp projections of 

 diatoms, especially the hollow spines of Chatoceros, it may yet be 

 that they are a defence against Crustacea, &c., as well, and this 

 may be equally true of Ceratium. At all events, I failed to find 

 evidence of Ceratiion being eaten, though it was diligently sought 

 for ; and repeated observations and estimates appeared to point to 

 there being a balance between the quantities of diatoms and animals, 

 irrespective of Ceratium. 



Reproduction of Diatoms. — In the ordinary process of repro- 

 duction, a diatom-cell divides into two cells, which resemble the 

 parent; and, in most cases, tlie successive generations remain 

 attached to each other in chains of individuals. Since the division 

 takes place at the girdle where one-half of the wall overlaps the 

 other, and each succeeding generation is similarly overlapped, and 

 since (unless, perhaps, in a free state) there is no subsequent 

 superficial growth, each succeeding generation is diminished in size 



* The late Dr. E. Brown in a paper " On the Nature of the Discolouration 

 of the Arctic Seas " (J'rans. Hot. Soc. Edin., vol. ix. p. 2i4, 1868), was the first, 

 so far as I know, to point out that this discolouration is due to the presence of 

 large masses of diatoms. He found these diatoms within the Pteropoda, 

 Medusce, and Entomostraca that compose the "whale's food." 



With reference to the association of diatoms with the food of fishes, Prof. 

 M'Intosh refers me to the Seventh Annual Report of the Board, in which he 

 anticipates my observations. At p. 272 (part iii.) he notes the occurrence of 

 Appendicularians with lihizo-sulenia. He says, " The stomachs of these were 

 distended by chlorophyll granules, the same appearing in the fa>cal pellets, both 

 in the intestine and when discharged." I regret that his observation was un- 

 known to me, and that I therefore did not cite it in my paper in Proc. Roy. 

 Soc. Edin. The main point, however, is the full and complete establishment 

 of the truth ; and, since my observations were made in entire ignorance of his, 

 this confirmation is all the more satisfactory. 



