468 THIRD PART. — BACTERIA OR SCEIZOMFCETES. 



refringentj splits across the middle in germination after its first distension into two 

 portions, which however remain firmly united at one side. The protoplasm invested 

 with a delicate membrane elongates in the direction of the longer axis of the spore 

 and of the longer axis of the mother-cell which coincides with it, and at the same time 

 usually bends through about 90°, and one extremity is thus thrust out of the opening in 

 the outer wall of the spore while the other extremity remains in the spore. The 

 protoplasmic body then developes into a rod. It appears from these phenomena as if 

 the direction of longitudinal growth in this case in germination is at right angles to 

 that of the parent-filament, but this is not really so. The wall of the spore which has 

 opened on one side is evidently very elastic ; it manifests so considerable a resistance 

 to the elongating rod as it bends, that the rod with both its extremities fixed in the 

 spore becomes curved before the one extremity is set free. The resistance sometimes 

 goes so far that both extremities remain in the spore, and in that case the elongating 

 rod assumes the form of a horse-shoe, the limbs of which may be of considerable 

 length. In other respects its growth is entirely the same as the ordinary growth which 

 begins with the formation of a bend in the protoplasm, and if the rod subsequently 

 divides into partial rods, two separate rods answering to the two limbs of the horse- 

 shoe may often be seen to project side by side from the ruptured wall of the spore 

 (Fig. 195 B, 4, 5). 



The cells produced in germination grow and divide by the formation of transverse 

 walls ; the products of this growth do not however remain united in a filament, but 

 separate one after another into rods consisting of a few or in many cases of a single 

 cylindrical cell, which may be 4-5 times longer than broad. These rods, which multi- 

 ply very rapidly and abundantly in fluids containing a good supply of food, display 

 active swarming movements all the time of the kind described above. I was unable to 

 satisfy myself with regard to the presence of cilia or flagella during the swarming stage 

 even in Bacillus subtilis. The rods become disseminated through the fluid as they 

 swarm and render it turbid. The last stage of vegetation is indicated by the entrance 

 into a state of rest, in which the cells increasing greatly in size remain united 

 together in filaments and the filaments form pellicles, till at last formation of spores 

 begins afresh. 



b. ARTHROSPOROTJS BACTERIA. 



Section CXXXI. In the course of the development of the species in this group 

 single members may simply separate from their connection with others, and under 

 suitable conditions become the initial members of new combinations ; they may there- 

 fore claim to be called spores. In other respects there is no general characteristic 

 distinction between them and the purely vegetative members. 



In connection with the fact that the species in this group are less like one another 

 than the Endosporous Bacteria, and that some species have a greater variety of 

 growth-forms, the mode of formation of the cells which may be termed spores varies 

 greatly in the different species. 



It is to the forms of this group that I give the name Arlhrobacterium proposed 

 for them in my introduction on page 454; they resemble Bacillus in their vegetative 

 form, and are usually known as species of Bacterium, but require a more exact 



