D. L. MACKINNON 
267 
are torn apart: by now they are usually so nearly equal in length that 
their independent movements tend to separate the daughter flagellates 
quickly from one another. This is the condition of things in the 
parasite from the fly. In that from the larva, as I have said, longitudinal 
splitting of the body often takes place before the new flagellum has 
grown up very far (text-fig. 1). The consequence is that the two 
daughter flagellates tend to hang together by their ends until such time 
as the new flagellum has grown sufficiently long and strong to work 
against the old one and so effect the separation. It was only by 
studying these processes in the living organisms that I came to under¬ 
stand the apparent rarity of the final stages of division in the fly, and 
their great abundance in the larva. 
After two or three hours the movements of the flagellates have 
become very much slower, and it can be seen that the body is shorter 
and more rounded, and has lost much of its rigidity. The flagellum has 
also become shorter, partly through actual loss of length, but chiefly 
through being drawn into the body, where it can be traced as a refractive 
line running back to the kineto-nucleus, now situated near the anterior 
border of the tropho-nucleus (text-fig. 2). This withdrawal of the 
flagellum continues until a mere stump is left protruding, and the 
kineto-nucleus is generally lost to sight among the refractive granules 
in the hinder part of the now pear-shaped cell. All forward progression 
of the organism gradually ceases, but the flagellar stump continues to 
move jerkily to and fro for hours. I have seen division take place at 
this stage, the longitudinal split starting posteriorly and running 
