1118 Proceedings of Royal Society of Edinburgh. [sess. 
stage of the auxospores, I am unable to say anything of the 
further development ; worthy of notice is the fact that they appear 
much later in the year than the auxospores of the other species. 
A phenomenon in the plankton Melosirce , which is rather inter- 
esting, is that the chains are curved , generally so much so that they 
form a spiral or corkscrew. In my note on phytoplankton from 
an Icelandic lake (1904, p. 233), I have mentioned the same pheno- 
menon from a lake at Heidi in Myrdalur (South Iceland), and have 
given two drawings to illustrate it (figs. 4 and 5). The species 
which is so curved has been defined by me as M. granulata , f. 
curvata , Grun., but Dr 0. Muller, who has seen the plankton 
sample, tells me that it is better to name it M. italica. 
In Thingvallavatn the M. italica is straight or only a little 
curved, while the M. islandica is often curved in the same manner as 
the form from Heidi. My drawings of parts of auxospore-bearing 
chains will give an idea of the degree of curvature ; the longer 
chains form often a corkscrew. There is no doubt that this cur- 
vature, as pointed out by me (1904, p. 233), must be regarded as 
an adaptation to the limnetic existence ; it is evident that the 
power of floating must thus be augmented. It is curious that the 
curved Melosirce are so seldom mentioned in the plankton litera- 
ture, although the Melosirce are very commonly found in the 
plankton of lakes in many different countries, and generally play a 
predominant part in the composition of the plankton where they 
occur. In a paper by E. Lemmermann (1904 b, p. 17) we find the 
following remarks concerning the evolution of new plankton forms : 
“ As the first influential factor, I take, naturally, the movement of 
the water ; it causes, e.g., the different curvatures of the Melosira 
forms. It is very rare to find quite straight chains in the plankton ; 
generally the stronger of them are more or less semicircularly 
curved, the weaker ones often being spirally twisted” (cf. also 
E. Lemmermann, 1903, p. 92). If we add to this that H. Yolk 
(1903, p. 133) in a table notes a curved Melosira as rarely occurring 
in a single sample, I have mentioned all which I have succeeded 
in finding on that question.* 
* O. Muller (1895) has, according to information given me by Mr Lemmer- 
mann, mentioned the curved Melosirce, but the paper has not been accessible 
to me. 
