THE LEUCO-THIOBACTERIA 95 



Tlie filaments frequently suffer dissolution with dramatic 

 suddenness, for the whole procedure which is detailed below 

 may occupy not more than a few minutes. The procedure 

 in autolysis follows one of two courses. 



First Method. — After coming to rest the filament breaks into 

 fragments, as shown in Fig. 6c. The sheath at this stage can 

 be made visible by treatment with iodine. Between the short 

 lengths into which the thread has divided (Fig. 6d) transverse 

 sheaths of slime are formed. Then, beginning with one of the 

 cells at or near the middle, each swells in turn (Fig. 6e). The 

 order of swelling seems to be strictly maintained. By the 

 time the fourth or fifth has begun to swell, the starting cell 

 has completely disappeared. The process continues until the 

 last cell has dissolved. 



Second Method. — When autolysis takes place in a filament 

 in which a well-developed and hardened sheath has already 

 been formed, a change in some of the details of the process is 

 observed. As before, the thread breaks up into a number of 

 short lengths, but the fragmentation is accomplished inside 

 the hardened sheath (Fig. 6h) ; and it is within the sheath that 

 the dissolution of the cells is effected. In Fig. 6h is shown an 

 example in which all except three cells have disappeared. 

 In this figure a septum of slime is also shown. 



Cause of Autolysis. — Winogradsky (2) states that autolysis 

 takes place as a result of the depri\-ation of sulphuretted hydro- 

 gen, but he did not advance any proof in support of this state- 

 ment. Actually, it takes place when inimical conditions of 

 growth reach a certain stage of intensity, and this may result 

 not only from the dearth of sulphuretted hydrogen but also 

 from an insufficiency or an excess of oxygen, or from other 

 causes. When the conditions of growth are not satisfied the 

 filaments disappear, so that in most cases not one but several 

 factors are concerned. It has been suggested that the process 

 is furthered by the secretion of a specific ferment, but no proof 

 of the existence of such a ferment has been advanced. It 

 is very unlikely that a process which is accompanied by a 

 series of remarkable morphological changes originates or is 

 furthered by a ferment. It must be borne in mind that the 



