144 W. M. Bale: 



close an ally of H. sujjerba that there is every probability that its 

 ramification would be similar; that is to say, a true branching, not 

 a dichotomous division. A definite distinction is seen in the 

 arrangement of the hydrocladia; those of //. arcuata are borne each 

 on a separate internode, and are consequently always alternate, while 

 in H. ascidioides every internode bears two hydrocladia, which are 

 always opposite or nearly so. The hydrothecae of H . arcuata differ 

 in form from those of H . ascidioides, being narrower towards the 

 base, and are more erect, the central axis of the latter being at 

 about 60 deg. to the internode. while that of the former is about 

 80 deg., the anterior side being indeed almost or quite at a right 

 angle. The hydropore in //. arcuata is more conspicuous than in 

 most species, owing to the perjsarc being thickened up to the edge 

 of the pore, so that the abrupt interruption is very noticeable 

 in optical section. The slight ridge or projection inside the 

 apocauline wall of the hydrotheca is caused by this thickened edge 

 being more or less incurved just where it borders the hydropore on 

 the upper side, and is therefore not analogous to the intrathecal 

 ridge in H. ascidioides (and all the members of its group), which 

 is a distinct septum, springing from the wall of the hydrotheca and 

 projecting half across its cavity, and which does not border the 

 hydropore but is situated some distance above it. Such a ridge i& 

 truly " intrathecal," while the ridge in H. arcuata is not within 

 the hydrotheca at all, but is merely a portion of its boundary. At 

 the same time there is no doubt that this rudimentary ridge in- 

 dicates how the fully-developed ridge of such species as H. 

 ascidioides has originated. The inflection of the hydrotheca-wall 

 has been extended till it reached half across the cell, and then the- 

 inflected portions have been brought into contact and united. But 

 this extension could not occur in a form like H. arcuata, unless 

 the hydropore were removed away from the ridge to a lower posi- 

 tion, as in the other species. While in most members of the genus 

 the hydropore is muclj less conspicuous than in H. arcuata, its 

 whereabouts is easily discernible (where the specimen is clean) by 

 the little points of perisarc which project from its upper and lower 

 margins. These denticles are a c-haracter which I have found 

 common to all the species of IlaUcornaria which I have yet observed. 

 In //. arcuata they are less conspicuous tlian in most species, but 

 one or two can generally be made out, at least on the upper border 

 of the hydropore, which, in this species, is the " ridge." 



In regard to the hydrotheca-margin the two species differ notably. 

 Both have an anterior tooth and one on the back, but the latter in. 



