318 II. JAMES-CLARK ON THE AFFINITIES OF SPONGES. 



In another minute or two the anterior indentation (fig. 16, e 1 ) had become so deep and 

 broad that the body presented a cordate outline when seen from its broader aspect, whilst 

 the furrow (e) appeared to extend to its base, and the distal end of the collar (b) had so 

 nearly closed up as to give that body an almost complete conical form, with a slightly 

 collapsed periphery. 



From this moment the process of reduction ceased, and soon after the cone-shaped collar 

 began to expand (fig. 17, b). Consentaneously with this, the anterior indentation (e 1 ) had 

 become sharper and deeper, and, — with the lateral median furrow (c) of each of the op- 

 posing broad flanks of the infusorian acting in combination with it, — had split the body 

 about half way to its base. The most remarkable phenomenon observable at this time was 

 what occurred at the rounded ends of the two half separated bodies of the new pair of 

 individuals. This was no more nor less than the incipient development of the flagellum, 

 which proceeded in this wise. At each of the rounded ends just mentioned a slight commo- 

 tion appeared, resembling the molecular vibrations of a granule, and then there arose quite 

 rapidly a sharp and distinct filamentous outgrowth (JI), which kept itself in a constant state 

 of narrow vibrations, or a sort of shivering. 



By 1.23 p. m. the newly born flagella (fig. 18, fl) had risen to half the height of the 

 collar (/;). and still remained in a shivering condition, whilst the body had divided almost to 

 its base, and the collar had broadened to a widely terminating, truncate cone. 



In about a minute more the dividing process had risen into the collar and split it 

 (fig. 19, e 2 ) upwards for one quarter of its height, and the still tremulous flagella (fl) were 

 slightly longer than in the last phase. 



By 1.26 p. m. the body was divided (fig. 20) to its posterior termination, and the fissura- 

 tion (e 2 ) of the collar (b) had reached half way to the distal edge, and was further sketched 

 out, as it were, by two opposing shallow, longitudinal furrows which extended to the margin. 

 At this period the collar was broader at the still undivided portion than below, so that on 

 the whole it had a very wide campanuliform shape, or rather, — since the divided portion 

 was rolled inwards at the opposing edges, — was like two slightly flaring broad funnels, 

 merged into each other at their broader ends. The flagella (fl) also had developed con- 

 siderably, and extended a short distance beyond the collar ; and the front end of the body, 

 from the middle of which the flagellum arises, had assumed the low, truncate, conical shape 

 of the adult form. 



From this time onward the division did not appear to go forward so rapidly, and the new 

 bodies seemed to be more particularly occupied in shaping themselves into the characteristic 

 form of the adult. The collar, however, was not long in dividing itself up to its margin 

 (fig. 21), but still the two cylindrical halves (b,b) did not separate at their extremes as soon 

 as the fissation reached that point. 



At 1.35 p. M. the self-division was completed (fig. 22) as far as the body proper was con- 

 cerned, and had extended a short way down the pedicel (pd 2 )- The margins of the two col- 

 lars (b) seemed merely to lie in contact, and each collar had a slightly funnel-shaped outline, 

 and was considerably more elevated in proportion to its diameter than in the adult form. 

 The flagellum (fl) was nearly as long as that of the full-grown body, but yet had neither 

 the sigmoid curve of the latter nor its stout and rigid aspect, but was much more delicate, 

 and in fact still exhibited a slight, tremulous motion. The two contractile vesicles (cv) of 

 each body were as distinct as those of the adult, and had the same proportionate size and 

 relative position. 



