SPIROCHATA BALBIANII AND SPIROCHATA ANODONTA. 43 
into several strips, and thus simulate a longitudinal division 
by folding of the membrane. In his own words: ‘Cette 
bande peut se diviser en plusieurs cordons et donne ainsi 
Villusion d’une division longitudinale de la ‘membrane ondu- 
lante.’” He.has probably mistaken the myoneme fibrils in 
this case. There is no illusion in the matter, for longitudinal 
division can be observed under the microscope, and so is 
undoubted. But Swellengrebel only sees one basal granule 
in 8. balbianii, while he maintains that there are two in 
Spirillum giganteum.- 
Trext-rig. 10.—a. 8S. balbianii; 4. S. anodonte, showing 
longitudinal division, The partially separated forms are gradually 
increasing the angle of divergence between them. A vacuole is seen 
in the cytoplasm at the undivided end. 
I can confirm in the main the details of longitudinal 
division as given by Perrin. 
The basal granules of each end appear to divide simul- 
taneously, but only those of the one end break contact 
and definitely separate, so that the daughter individuals 
resulting from fission remain attached at one end often for a 
long time. Occasionally, however, | have noticed forms 
attached by each end, as figured by Lustrac (6, Pl. 1, fig. 1). 
I think such were true cases of longitudinal division, and 
not mere intertwining of two forms lying side by side, but 
otherwise unrelated. 
