1116 Proceedings of Royal Society of Edinburgh. [sess. 
sent some samples and slides to Dr Otto Muller of Berlin, who 
has worked especially with this genus (1903, 1904). He has been 
so kind as to name the coarser species M. islandica , O. Muller, 
n. sp., and the narrower one “ M. italica, Ktitz., f. tenuis and 
f. tenuissima , with forms which pass into M. crenulata, Ktitz.” 
The M. islandica is most nearly allied to M. granulata, from 
which, according to his letter, it differs in the shape of the pores, 
in the absence of marginal teeth of the valves, and in the form of 
the so-called “ mutation.” Dr Muller will examine the species 
more thoroughly and publish the results later on in a separate 
paper. In a letter he tells me that the interesting facts of the 
different development of the pores in the two halves of a cell, or 
in the different cells of a chain, as well as in the different chains 
of a sample — a fact which he explains as mutation in the sense 
of H. de Yries — also occurs in the species from Thingvallavatn. 
He has given an interesting paper on this subject (1903), and he 
will now take also the two Icelandic forms into the examination. 
Dr Wesenberg-Lund and I are very happy to have obtained help 
from such an authority with regard to Diatoms as Dr Muller, 
and we desire to express to him, here, our best thanks for his 
kindness. 
Dr Midler will probably treat the question of the auxospore forma- 
tion from a more systematical point of view, but here I only intend 
to give a description of the development of this propagation method. 
In M. islandica I have found numerous chains wdtli auxospores, 
and thus I have been able to follow their formation rather well. 
The drawings in PI. I. figs. 1-7 will show some of the successive 
stages. The formation begins in the same manner as the ordinary 
cell-division : a cell in the chain prolongs itself by moving the 
connecting parts apart from each other, the plasma becomes con- 
centrated in the new-formed, thin- walled part of the cell, swells up 
and forms a sphere. At that time the cell bursts, because the 
cohesion between its two halves is diminished by the formation of 
the globular body ; we therefore always find the auxospore at the 
end of a chain, and often the chains have an auxospore at each 
end. At first the auxospore has no siliceous wall ; it increases 
until it reaches it legal dimension, then it produces a rather thick 
siliceous wall with a very low connecting part consisting of one 
