PHYSOPHORA. 79 



the air-vesicle " has the form of a pear, and is enclosed in a fibro-flocculent tissue, which lines 

 the interior of the pneumatophore {partie pyriforme), surrounds the pigment-spot, and 

 descends, forming four organs, in the midst of which passes down a rounded mass of 

 the same substance, and closely invests the air-bubble, which thus perfectly resembles the 

 clapper suspended in the middle of a bell." 



I found the structure of the pneumatophore to be quite similar to that which obtains in 

 Agalma; that is to say, the air is contained in a distinct pneumatocyst, enclosed within 

 a reflection of the endoderm of the pneumatophore. No cellular processes, like those of 

 Hhizophysa, were observed to proceed from the under part of the reflected endoderm, 

 and I could discover no pore communicating with the exterior, even in the youngest 

 specimens. 



The coenosarc, immediately below the pneumatophore, gives attachment to a number of 

 budding nectocalyces, which are succeeded by a double series of perfect ones. Of these 

 there may be as many as five on each side.^ ' 



Kolliker says that these are so like those of his Agalmopsis, that it is unnecessary 

 to describe their form particularly ; they alternate and embrace the coenosarc by their 

 excavated sides. This was also the case in Vogt's specimens and in mine. According to 

 the last-named investigator, the nectocalyces are horseshoe shaped, with an orbicular aperture 

 on the truncated summit of their external curve, surrounded by a membranous valve, 

 and leading into a nectosac, which is " very small, and does not extend through more than 

 half the thickness of the organ." 



M. Vogt's figures, however, represent the nectosac with much larger relative proportions, 

 and Kolliker represents the nectosac with much the same proportions as in my specimen, 

 in which the nectocalyces were about a quarter of an inch long, and had an irregularly 

 cordate form. Each was attached by a broad, thick pedicle to the coenosarc, and its 

 axis was inclined (as Vogt notices) at an angle of 45° to the latter. The inferior and 

 external face of the organ was broad and convex above, but suddenly narrowed towards 

 its free end. The superior and internal face was deeply excavated in the middle line, so 

 as to present a convex lobe on each side, between which arose the pedicle of attachment. 

 The excavation was continued into the internal and inferior face, while the inferior and 

 external face was truncated and almost wholly occupied by the rounded, or somewhat 

 four-sided, aperture of the nectosac, provided with its narrow membranous valve. 



The edge of junction of the inferior internal, and inferior external, faces is prolonged 

 into a thin, broad, subquadrate plate, whose lower free edge is emarginate in the middle line, 

 and pointed at the angles. 



The sides of this plate are continued into two ridges, which run along the margins 

 of the inferior internal face to about its middle, and then become constricted, but are 

 continued upwards, on each side, on to the superior internal face, and thence downwards 

 on the superior external face, as far as its truncated inferior edge, where they end in 

 obtuse points. Superiorly and internally these ridges form broad plates, which overlap 



^ In P. tetrastkha Philippi found four series of nectocalyces, each series containing four 

 of tliese organs (p. 59). It would appear, nevertheless, that they were all really attached to one side 

 of the coenosarc. 



