248 Transactiuns of the 



accompanying flagellum could not yet be made out. But in four 

 hours from this time the field swarmed with active monads, spring- 

 ing like the parent forms as seen at Fig. 5. 



The way in which the flagella first appear in these germinal 

 forms w-e have not discovered : but the proboscis flagellum appears 

 to lengthen by motion up to a certain point ; and we have satisfied 

 ourselves that the proboscis is extruded sarcode, being the earliest 

 diflerentiation of the granule. The nucleus first appears as a black 

 speck and then slowly enlarges. 



It remained then for us to discover the relation between the 

 form 1, PL XLIIL, and the ordinary form, Fig. 6, PL XLI. 



We did this by constantly watching the behaviour of the forms 

 we had seen germinally develop. 



In the vast majority of cases nothing but the fission first de- 

 scribed and figured in PL XLI. was seen, and it would appear that 

 in the " grooving cell " at least, that this process is exhaustive, con- 

 tinuing only in vigour for a certain time, and then becoming 

 weaker, and at last ending m death. 



But amongst the mass of anchored forms some few were seen 

 (much larger than the others) which occasionally detached them- 

 selves and swam slowly, the trailing flagellum flowing gracefully 

 behind ; as shown in Fig. 5, PL XLII. It was watched with -V and 

 No. 2 eye-piece. In the course of an hour it became still ; but both 

 its flagella were free. An amoeboid condition supervened, causing 

 the whole substance of the sarcode to be pushed out to a and h, 

 Fig. 6, while a large disk c is constantly present in this stage, and 

 exhibits an opening and shutting motion like that of the eye-lid, 

 opening at either hand from a median line and snajpjping with great 

 force. In the course of three hours it had passed from an oblong 

 into a rough lozenge -shape, and from that to a disk, the flagella being 

 still i^attached and waving. Fig. 7. The vacuoles d, e, gradually 

 condensed into a dark globule, and a small cone of sarcode was 

 pushed out at c, while a line fi-om a to 6 became shortly visible. 

 What followed this was simply a repetition of what is recorded on 

 a preceding page, and is drawn in Figs. 1 to 8, PL XLIII., and 

 Fig. 1, PL XLI. 



We had thus gathered up the threads and completed the life 

 history. The usual method of multiplication is by fission, which 

 goes on apparently to exhaustion. Amongst enormous numbers 

 there are a few distinguished from the others by a slight increase in 

 size and the power to swim freely. These become still ; — for a time 

 amoeboid — then round ; a small cone of sarcode shoots out, dividing 

 and increasing into another pair of flagella The disk splits — each 

 side becomes possessed of a nuclear body, and two well-formed 

 monads are set free. These swim freely until they attach them- 

 selves to an ordinary form that has just completed fission, so that 



