432 MB. A. W. BENNETT ON THE 
that when two lie close together, it is often very difficult to 
follow either of them with certainty throughout its whole length; 
and I have often myself been deceived at first in this way. On 
the other hand, De Bary (l. c. p. 4) says:—“ It is easy, with 
almost perfect certainty, to determine beforehand which of two 
filaments is the one that parts with, and which the one that takes 
up the protoplasm, if only conjugation has already taken place 
in one pair of cells ; for almost always all the cells of a filament 
behave in this respect alike." H. C. Wood does not refer to 
this point. Cleve*,in his numerous drawings, gives two instances 
of what we may call “ eross-conjugation ;” Cooke, in 11 plates, 
not one. If, therefore, the phenomenon occurs, it is probably 
very rare. In my own observations, extending over several years 
and many species of Spirogyra and Zygnema, I have never seen à 
single instance. In one example only I observed the protoplasm 
in the cell of a male filament collecting into the form of a zygo- 
spore in a part of the filament where communication with the 
female filament was interrupted ; but there was nothing to show 
that a fertile zygospore resulted. A further evidence of differ- 
entiation lies in the fact that, as far as my observation goes, when 
two filaments are about to conjugate, the change first commences 
in the cells of the male filament, the chlorophyll-bands of which 
lose their spiral arrangement, accompanied by contraction of the 
protoplasm, before those of the female filament (see fig. 2). 
The differentiation of the male and female filaments is even 
more striking when more than two are in conjugation, as is not 
unfrequently the case, and as has been repeatedly observed by 
De Bary and others (fig. 1). As far as my own observation goes, 
polygamy is here much more common than polyandry. I have seen 
different cells of the same male filament pass their contents into 
the cells which happened to be nearest them of no less than four 
female filaments. In one instance as many as thirty-six cells of 
the same filament passed their contents into other filaments, 
without transference in the opposite direction taking place in à 
single cell. It is very curious to watch a male lying between 
two female filaments and passing its cell-contents sometimes into 
one, sometimes into the other. Cleve (7. c. pl. v. fig. 8) and Cooke 
(l. c. pl. xxxv.) have figured the same. This tendency to poly- 
gamy appears to arise from the fact that the male are very muc 
* * Monografi af algfamiljen Zygnemaceze.' 
