1 882.] The Siphonophores. 93 



of Agalma. The apex of the anterior bell is in reality the pro- 

 longation of the side of the bell, while the true apex has been 

 abnormally twisted out of position, and is found just above the 

 bell opening, near the origin of the stem which seems to hang 

 down between the two nectocalyces. 



Nowhere in its structure is the modification, which takes place 

 in the organs of the bell as a result of this abnormal twisting, 

 better shown than in the course of the chymiferous tubes upon 

 the inner walls of the bell cavity. The radial tubes are especially 

 modified in their course by the change in external form which the 

 bell has undergone. 



The chymiferous tubes of the anterior nectocalyx in Diphyes, 

 consist of a system of four radial vessels placed upon the inner 

 walls of the bell cavity, and a single large tube or cavity extending 

 into one side of the bell walls parallel to the outer surface. The 

 former tubes start from a common junction, and pass radially to 

 the bell margin, while the latter ends blindly about two-thirds the 

 distance between the bell rim and the pointed extremity of the 

 nectocalyx. Both open into the cavity of the stem ; the former 

 by a vessel passing from their junction to the stem; the latter 

 more directly through the same tube. 



The length of the four radial tubes is very unequal, as would 

 naturally be expected if the distortion which we have suggested 

 as having taken place in the anterior bell, has in reality occurred. 

 The two tubes (c) which lie in those parts of the bell which have 

 been enlarged, are therefore naturally much longer than those in 

 the remaining portions of the bell. So small indeed has that 

 side of the nectocalyx which adjoins the posterior bell become, and 

 so enormously has the opposite half been enlarged, that the tubes of 

 one are inconspicuous and with difficulty traced, while those of 

 the other are very prominent on the inner bell walls. At first 

 sight then, we might suppose that there were but two radial ves- 

 sels, while a closer study shows that there are four such tubes as 

 we have seen exists in the nectocalyces of all Siphonophores. At 

 the common junction of these tubes, we must look for the 

 apex of the bell cavity. At that point, about midway in the 

 length of the two bells, the vessels communicate with the stem 

 cavity by means of a short tube, similarly placed to a like vessel 

 in the nectocalyx of Agalma. 



There is, however, in the anterior nectocalyx a tube (I) which 



