288 PROCEEDINGS OF SECTION D. 



curved to one side, the tips bend at right angles, and the other 

 branch grows out from the angle. Tn others {e.g., Cos. subcrenahim, 

 Hantzsch, Rein.) the spines are straight and rod-like with blunt 

 ends ; the ends divide evenly into two branches. It may, I think, 

 also be taken as a general rule that mature zygospores are spherical, 

 those which are elliptic, subglobose, or quadrate (with a fair 

 number of exceptions), being not yet perfectly formed. In cases 

 where the cells fuse together the zygospores are often more or less 

 cruciform at first ; nevertheless, they soon become subquadrate, 

 then subglobose, and finally spherical. On this point compare 

 Ralfs, T. XXV, fig. 6, where he gives a complete series illustrating 

 the formation of the zygospore of Cylindrocystis {Peniitm) Brebissonii 

 Cf. also De Bary, Conjugatae, T. v, figs. 26-27. 



Pyrenoids. — In the genus Cosmarhmi, forms of which have 

 specially engaged my attention lately, it used I believe to be con- 

 sidered that cells with double pyi-enoids were specifically distinct 

 from those with single ones. (See Delponte, Desm., suhalp., 

 p. 275, where the species are classified thus.) How far this is 

 accepted to-day I have no means of knowing. In a recent 

 gathering, however, I came across a cell of Cos. sitbspeciositm ji 

 validiiis, Nord., forma, in which just above the isthmus was a single 

 P3a-enoid, giving by its appearance every indication of age, while 

 above were two others, apparently much fresher and younger. 

 The impression could not be avoided that the single pyrenoid was 

 a legacy from the younger stages — the cell was even then not 

 nearly mature. A strong suspicion has formed in my mind there- 

 fore that probably in this genus only the variations with binate 

 pjn-enoids are approximately fully developed, that the perfect form 

 of the species of Cosmarium always bears geminate pyrenoids, and 

 that variations with only one to the semi- cell are more degenerate 

 forms. If this is true, two things result — first it does away with 

 all minute and small species (in the true sense), as these small forms 

 almost always have but single pyrenoids, and secondly it provides 

 a terminus ad quern towards which one may look in working out a 

 life-history, since we cannot have arrived at or near the end till 

 we have traced a connection with such a form. Under any cir- 

 cumstances I see very good reasons for believing that almost all 

 the smaller kinds of Desmids are merel}^ degenerate forms of larger 

 ones, and not distinct species at all. As I remarked at the outset, 

 in Desmid life degeneration and multiplication triumph, as a rule, 

 at the expense of individual perfection. We must not be surprised 

 under such circumstances to find in nature a multiplicity of imma- 

 ture forms and relatively few distinct species. 



In the case of Cos. siibspeciosum forma cited above, the old 

 pyrenoid had occupied no doubt the normal place for a single 

 amylaceous nucleus, viz., the centre of the semi-cell, from which it 

 had dropped to a position a little above the isthmus, and the two 

 new ones were in the usual position for double pyrenoids, viz., 

 a little to right and left of the centre. These latter did not look 



