292 PROCEEDINGS OF SECTION' D. 



increase in size in all directions at once, a thing that only takes place 

 in the early stages of cell-division. Desmids do grow, and in 

 growing they develope from one form to another ; it is entirely 

 erroneous to regard each form as fully developed and unchangeable. 

 But increase in size takes place according to certain definite rules ; 

 every species indeed has peculiarities of its own, but the general 

 rule applicable to all is that increase in size takes place alternately 

 in length and breadth. 



Granules, Etc. — Granules, verrucae, teeth, spines and pro- 

 cesses are simply rucks in the cell membrane produced by obscure 

 forces proceeding from the cytioplasm.. Their gi'owth would 

 appear to be always connected with the chloroplasts, and some- 

 times with an angle or ridge in the chloroplast. In Docidium, 

 where the chloroplasts are bent suddenly round at the apex, rugge, 

 teeth or even elaborate tubercles make their appearance exactly 

 opposite these angles ; and very probably the variation in the 

 number of tubercles in the forms of Doc. coronulatum is due to 

 some difference in the size or shape of the chloroplasts at this point. 

 A most striking case is Cos. magnificum. This Desmid is covered 

 with abruptly truncate verrucae, which viewed from above are seen 

 to be circular, with three granules in a triangle on each. Moreover, 

 occasionally these verrucae grow out into decided teeth (? Cos. 

 dentictdatum, Borge). It is not surprising therefore to find when 

 the cytioplasm has been shrunk with preservative that a separate 

 filament projects into each verruca. Cos. Askenasyi, Schm., has 

 its endochrome even more cut up into fibrils on the outer side, and 

 though it seems to have been originally smooth, yet it is obvious 

 that verrucae are making their appearance. 



Again, in those Statirastra which have a circle of processes at 

 the apex, these processes are always arranged above the six 

 radiating plates of the chloroplast and correspond in number to 

 them ; and if there are processes down the angles they are almost 

 always in pairs over the two chloroplasts. In Xan. bifurcatum, 

 Borge, and Xan. gloriosiim, G. S. West {X. pulcherrimum, Playf.) 

 there are six ridges in the chloroplast which correspond to the six 

 apical processes, and the lateral processes are also on the lines of 

 these ridges. Somewhat the same state of things is to be found in 

 Etiastrnm in Eit. dideltoides, Rac, at any rate the chloroplast 

 exhibits six longitudinal plications, connected, one with each side 

 of the semi-cell and the other four with the lateral pairs of infla- 

 tions. The lateral basal inflations, which are the most pronounced 

 in this species, lie directly over the sudden bend in these last, where 

 they twist round towards the centre of the isthmus. The rule seems 

 to be that inflations, processes, spines, teeth, granules, etc., follow 

 the ridges of the chloroplasts and have some direct connection 

 with them. It seems probable that even the puncta-gi'anules mav 

 be the outcome of some invisible difterentiation on the surface of 

 the chloroplast. 



