MICROPHYTES POUND IN THE BLOOD. 393 



blood mixed with fresh aqueous humour, and placed in the same 

 chamber. 



On the following morning this slide, to which the half per 

 cent, salt solution had been added, was re-examined, and it was 

 found that the filaments had grown greatly in length and some- 

 what in thickness (PI. XVII, fig. 10) ; in some instances the 

 filaments extended across the field of the microscope. All the 

 filaments were motionless and almost translucent, quite devoid of 

 granularity, and it was only in some places that a joint could be 

 distinguished. No refringent molecule appeared in any of these 

 long filaments, but there were some short, pale, transparent 

 rods rolling about in the preparation, and in these glistening- 

 bodies were found (PI. XVII, fig. 12). Some of these rods, or 

 segments, were 8 fi long and contained a bright blue (as seen with 

 Hartnack's No. 9 immersion objective) ' spore,' 2 n in length 

 by 1 ^ in width, and other segments, about the same length, 

 contained two. Mixed with these were short, translucent staves, 

 with a distinct joint, some with two ' spores,' separated by a 

 partition, and others shorter (-t*5 fi) with only one. By the next 

 day the filaments were broken down and the preparation consisted 

 chiefly of a multitude of active Bactermm termo. 



The other slide, which had been prepared with aqueous humour, 

 was likewise examined on the following day. The filaments were 

 not so long as in the other preparation, and there appeared to be 

 a decided tendency towards cleavage into small cuboid pellets of 

 plasma (PL XVII, fig. 11, a). Some of the filaments, though 

 well preserved at one end, were seen to be undergoing the 

 process of fission at the other, each fragment being equal to 

 1 — 1'2 n in its longest diameter. It seemed as if the 4 to 5 ^- 

 segments, of which the filaments were composed, had first become 

 freed from the thread, and had, instead of giving rise to a 

 'spore,' undergone fission (fig. 11, ^). In other cases cleavage 

 of this kind took place whilst the individual segments maintained 

 their linear arrangement (fig. 11, c). In some instances it 

 seemed as if the two first halves of the originally 4 to 5 /i-seg- 

 ments had each become elongated (and correspondingly thinner) 

 and undergone further division, thus forming four more or less 

 spherical plastides (fig. 11, d). When the whole filament had 

 undergone such a process and the plastides had retained their 

 linear arrangement, it presented the appearance of a rosary chain 

 (fig. 11, e). It was ascertained that four of the plastides form- 

 ing a part of the particular chain sketched were equal to the 

 length of one of the segments of the original filament, viz. 5 ju. 



It will thus be seen that filaments of bacilli may disappear at 

 least in two ways : (1) by giving rise to minute highly refractive, 

 long-oval molecules, the filaments themselves becoming at first 



