466 THIRD PART.—-BACTERIA OR SCHIZOMFCETES. 
The removal of the transversely-opened membrane is the best and only certain 
demonstration of the fact that the spore is actually invested with a special membrane. 
This membrane is easy of observation, especially in cultures in very flat drops of the 
fluid. It sometimes divides quite across into two halves, which are attached at first 
like caps to the extremities of the elongating rod and are afterwards torn from them. 
I often failed to observe the final removal of the membrane in larger quantities of the 
fluid. There each cap, distinguished by. its more 
sharply defined outline, was seen to be attached to 
the extremities of the growing spore and gradually 
ceased to be distinguishable ; it would seem therefore 
that in these cases the membrane, after splitting trans- 
versely, is not removed from the spore in connected 
pieces, but swells up or is dissolved and so disappears. 
The membrane often clings to the rod-cell which has 
slipped out of it in such a manner that the longer 
diameters of the two bodies cross one another, and 
hence the appearance is produced as ‘if the growth 
in length in germination is in a direction at right 
angles to the longer diameter of the spore or parent- 
rod; but this is not really the case. 
The species which we are considering occurs 
also in another form, namely in long curved chains, 
in which a segmentation of the smooth cylindrical 
rods is only obscurely distinguishable, while the 
isodiametric cells on the other hand are distinctly 
defined, are slightly swollen and protuberant, and are 

Fic. 195. A Bactllus Anthracis. 
Two filaments grown on a microscopic 
slide in a solution of meat-extract, partly 
in an advanced state of spore-formation ; 
at the top two ripe spores which have 
escaped. The spores are drawn some- 
what too narrow; they are nearly as 
broad as the transverse diameter of the 
mother-cel. B Bacıllus subtilis. 1. 
fragments ofa filament with ripe spores. 
2. spore beginning to germinate; the 
outer wall ruptured transversely. 3. 
young rod projecting from the wall of 
the Spore in the usual transverse 
position. 4. Germ-rods curved in a 
horse-shoe shape and with the extremi- 
ties connected; one of them with one 

Germ-tubes with the two extremities re- 
maining connected and already greatly 
increased in size. Magn. 600 times, 
in many cases themselves curved at the points of 
greatest curvature. Torulose chains of this kind or, 
to use the terminology noticed above and founded on 
growth, coccz grouped in rows, are usually developed 
in great abundance, being loosely intertwined into 
thick convoluted masses and sometimes breaking up 
here and there into separate cells. They are motionless 
or show a very slight movement. There was little 
or no formation of spores. I almost invariably dis- 
covered this form when the cultures were rendered 
impure by other small Bacterium-forms, but the 
Bacillus formed the majority. It may be a question 
how far this fact points to the cause of the appearance of the torulose chains, 
but it is certain that they become transformed into the even rods when the culture 
is pure. 
The cells of the Bacillus of anthrax, B. Anthracis, Cohn (Fig. 195 A), ap- 
proach 1 p in thickness in vigorous specimens grown in the blood of animals attacked 
by it, and grow to about 3-4 times that length. They are connected together in the 
blood into straight rods of different lengths; in cultures on some suitable dead sub- 
stance they[form long pluricellular filaments; these may be much twisted or bent into 
