110 OOCYSTIS AND EREMOSPH.ERA, 



latter. Similarly, if Chodat's figures of 0. lacustris be consulted 

 (Etudes de biologie lacustre, in Bull, de I'herbier Boissier, T. v., 

 Pl.lO), apiculate (not merely acuminate) forms may be noted, 

 approximating to 0. ajricidata (cf. Text-figs. 5, 13). A form, 

 then, may be characteristic of one "species," and yet biologically 

 connected with another. 



0. hexago7ta mihi, has also this mucous lining, and is biologi- 

 cally a variant of 0. apiculaia, the connecting link being 0. 

 apiddata var. obesa; it is often very difiicult to say exactly to 

 which of them a specimen belongs. 0. hexoffona, however, is of 

 so distinct a shape, and so widespread, that it has seemed best 

 to make it a new type. 



0. c7-assa, 0. lacustris, and 0. ^jarva seem to me to form but 

 one species; they are only very slightly different, and are all 

 plankton-forms. 0. australieiisis mihi shows the same shape, 

 repeating itself, but in a swamp-form. 



Again, 0. ovalis, though its chloroplasts are of a type found in 

 no other species, yet reproduces, on a large scale, the same set of 

 forms as are found in 0. Nagelii {elliptica West). 



And further, the figui'es in Bull, de I'herbier Boissier (on 0. 

 Iacust7'is), and in Wille's study of 0. siibniarina, show that, in 

 both these species, the smaller, reduced forms are 0. Novce Semlice 

 pure and simple. 



The nearest approach to a fixed type is, perhaps, 0. solitaria 

 Witti'., of which only a f. niajvr and var. maxima have been 

 noted, these being merely a reproduction of the type, with greater 

 dimensions. Yet even so, 0. apiculata has a form (var. majvr 

 mihi) which counterfeits 0. solitaria very closely.* Indeed, one 

 might say that we keep our types distinct by arranging inter- 

 mediate forms as variations. 



Just as reduction-forms are produced in the Desmidiacece 

 by rapidly repeated subdivision, so also, in the Antosporacea', 

 they are brought about by the rapidly repeated formation of 



* O. solitaria var. notabile West, .also approximates very closelj^ in out- 

 line to 0. apiculata. c 



