IOO 



the sigmoid curve strongly marked, .with a branch arising from each convexity. In the first 

 internode shown, there are 24 zooecia, a branch originating on the right side from the fourth 

 zooecium of that side, and a second branch much higher from the tenth zooecium of the left 

 side. At about the middle of its length the fertile internode bears an ovicell, which is the 

 moditïed fifth zooecium of the left or inner side. The zooecium which immediately follows the 

 ovicell, on the same side, is displaced by the growth of the ovicell, as is usual in Crisia ; and 

 has a specially long peristome, which is seen, on the left margin of the internode, just distally 

 to the ovicell. The peristome in question is distinctly narrower than the ordinary peristomes; 

 a feature which I have found commonly, but not universally, in the present species. The fertile 

 internode has the arrangement which I regard as typical of such internodes in the present 

 species, and as distinguishing it from C. denticulata ; a proximal branch being developed from 

 the outer side of the internode, and two others, on either side, from the region of the distal 

 part of the ovicell. 



Using the notation which I have employed in my previous paper, the part of the zoarium 

 shown in fig. 1 may be represented by the formula : ) : — 



(a) Slide 107. B. 2 — (2 4 + r 4 + io'j + (iS + r 4 + S' +*) 



I 

 (9 + Ov. -f 15 + r 3 + r. + y + x) 



the full number of zooecia being stated for the second main internode, which is only partially 

 represented in the figure. 



The colony of which a part is here shown is remarkable for the facts ; — that most 

 of the internodes are of great length, the number of zooecia varying from 22 to 32; that most 

 of them have an even number of zooecia; and that they usually give rise to two branches. 



The examination of other specimens shows, however, that these relations are by 110 

 means universal. This is brought out by the following formulae : 



(b) Slide 107. R. 2 (another branch) — (19 + 7') -f (21 + r.) + (19 + 7') + (19 + ;-,) 



(c) Slide 107. B. 2 (another branch) — (il + r 5 ) -f (24 + qr + /-,„) + (15 + 5') 



(d) Slide 313. A — (s + 2 r) + (7 + rJ + (9 + 4r) + (9 + r,) + (9 + 2r) + (11 + rj 



I 

 (11 + 2 '- + r 3 ) + 



In these formulae most of the internodes have a single branch and an odd number of 

 zooecia; but in [c] one of them agrees with (a) in having two branches and an even number. 

 This seems to be the general rule, but it is not universal, as is shown by (d), where an odd- 

 numbered internode bears two branches. 



When only a single branch occurs, its position varies, usually from the 2 nd to the 7 th 

 zooecium of its own side. The formula (d) represents a main stem, attached by rootlets at its 

 base. It illustrates the tendency which has already been pointed out of forming at first short 

 internodes with few zooecia and the branch arising as low as the 2 nd zooecium. These small 



1) The limits of the internodes are indicated by the brackets. The numerals denote the number of the zooecia in an internode: 

 Ov = ovicell: x = growing point: r t = a branch given off by the 4 th right zooecium: 10' = a branch from the tenth left zooecium. 

 Juli.ikn and Cai.VET (1903, p. 109) consider this a useless rnethod of notation; but it appears to me a convenient method of recording 

 some of the characteristic features of the species. 



IOO 



