DIATOMACE 195 
emerge from the valves and unite to form one body, 
which then grows into an auxospore. There seems 
to be some doubt whether there is here a true con- 
jugation—whether these two cells are true motionless 
gametes and the product azygote. (4) This process 
may be gone through, but without.union. The two 
cells give up their contents, which lie side by side 
unclothed with a membrane, and either close together 
or separated by a layer of the gelatinous envelope. 
Each develops independently into an auxospore. 
(5) Two auxospores may be formed by a pair of 
Diatoms, which on emerging from the old valves 
divide each into two. Each pair of the four cells 
thus formed conjugate and form an auxospore. Re- 
garding all these processes, it appears to be fair to 
assume that conjugation, or rather union, whether 
between one pair or two pairs, is the normal process 
of forming an auxospore, and that the other cases 
are parthenogenetic, where a pair produce auxospores 
together without conjugation, or where one produces 
two auxospores by itself, or where one similarly 
produces a single auxospore. The simplest form of 
all is common ; so is the case of two Diatoms forming 
their auxospores without conjugation, and of two 
combining to form one by conjugation, though not 
so frequent as the others. The case of one Diatom 
forming two auxospores is known in only one form, 
viz. Rhabdonema arcuatum ; and the last case, where 
two Diatoms divide and the daughter-cells conjugate, 
each pair giving rise to an auxospore, is recorded for 
Epithemia Zebra only. The mode of formation of 
Oo 2 
