126 DR A. MILNES MARSHALL ON THE 



the largest attain a length of 1*6 mm., with a width of 13 mm. In transverse 

 section (figs. 10-16, i) the spicules are triangular, with rounded angles, and a 

 shallow groove running somewhat obliquely down the middle of each face. 

 They are exceedingly numerous along the whole length of the abaxial surface 

 of the zooid, and are arranged with their long axes parallel, or nearly so, to 

 that of the zooid. 



The mesoderm is traversed by a system of irregularly branching nutrient 

 canals continuous with those of the rachis. The muscular system of the body 

 wall of the zooid seems to be completely absent. 



The relations of the inter-septal chambers in the part of the zooid above the 

 stomodreum are rather curious. Fig. 14 represents a transverse section through 

 the mouth opening ; it shows that at this point only five of the eight inter- 

 septal chambers are present, viz., the abaxial, or as it is commonly called, ventral 

 cavity, the two latero-ventral cavities bounding it on either side, and the two 

 lateral cavities ; the axial or dorsal and the two latero-dorsal cavities not 

 ex-tending above the mouth. In a section taken a little higher up, through 

 the upper part of the mouth (fig. 13), the two lateral cavities have dis- 

 appeared, and the mid-ventral and latero-ventral cavities are alone present. 

 Tracing them further up towards the apex of the zooid, we find that all three 

 persist for some distance, but that after a time the middle abaxial or ventral 

 cavity, which has been from the start the smallest (cf. figs. 15 and 14), loses 

 its lumen (fig. 12), and then disappears altogether, the two latero-ventral 

 cavities alone persisting (fig. 11). 



Further up still (fig. 10), one of the two latero-ventral cavities disappears 

 and one alone is left, which can be traced nearly, or in some cases quite, up to 

 the apex of the zooid. 



The prolongations upwards of the interseptal chambers above the mouth 

 correspond, not to the tentacular cavities of the fully formed polyps, but to 

 the cavities of the calyx processes ; and the whole of the part of the large 

 zooid above the mouth is to be regarded as formed by a special unilateral 

 development of the calyx, corresponding at its base to five, and along the 

 greater part of its length to three calyx processes fused together, but with 

 their axial cavities remaining distinct. 



That the spine of the large zooid really consists of calyx processes and not 

 of tentacles, is, I think, proved by the perfect continuity between the wall of 

 the zooid itself and the spine ; by the unbroken series of exceptionally large 

 spicules extending along the abaxial wall of the whole length of the zooid, 

 including the spine ; and by the absolute identity between a transverse section 

 across the upper part of the spine, and one through a calyx process of a normal 

 polyp. This latter point is well shown in figs. 9 and 10, the former being a 

 section of a calyx process, and the latter of the spine of one of the large zooids. 



