vol. 2, pt. 2.] A TAXONOMIC STUDY OF THE SALPIDAE METCALF. 39 



Opposite the lower part of the intermediate muscle are a pair of 

 languet-like processes of the mantle, one on each side. They resem- 

 ble the corresponding structures in the solitary Cyclosalpa pinnata. 



The ganglion, the neural glands, the outgrowths from the ganglion 

 toward the neural glands, and the horseshoe-shaped dorsal eye are 

 as in the solitary forms of the other Cyclosalpas already described. 



CYCLOSALPA BAKERI, aggregated form. 



Plates 8, 9, and 10. 



In the aggregated form of this species we find again considerable 

 resemblance to the aggregated Cyclosalpa jloridana, in general shape 

 and in the very flabby character of the animals, but they are sharply 

 distinguished by many features, as Ritter and especially Ihle have 

 shown. The younger zooids, 4 mm. long, just detaching from the 

 stolon, are different in shape from the larger Philippine specimens, 

 12 mm. long. The older individuals (pi. 8, fig. 21) have a longer 

 atrial siphon and they seem to have lost the humpbacked form so 

 noticeable in the younger zooids, but our three older individuals are 

 all so soft and flabby that it is difficult to be confident of the exact 

 shape. It is, however, certain that the dorso-ventral diameter in 

 front of the atrial siphon is relatively greater in the younger indi- 

 viduals. A marked difference between the older and younger zooids 

 in the arrangement of the oral muscles will be noted when we reach 

 this part of the description. 



The body muscles have a different arrangement on the right and 

 left sides of the body, this species being one of the Cyclosalpae asym- 

 metricales. The intermediate muscles and their prolongation into 

 the peduncle are as in the aggregated Cyclosalpa Jloridana. The 

 first and second body muscles on the left side are fused dorsally and 

 ventrally, but are distinct and widely separated through the middle 

 part of their course. Both share in the formation of the posterior 

 peduncle muscle. On the right side of the body there is a diagonal 

 strand which we interpret as the second body muscle fused at its 

 lower end with the first body muscle, and in its upper portion fused 

 with the third body muscle. Comparison with the aggregated Cy- 

 closalpa Jloridana suggests that, in the latter species, the muscle we 

 have numbered I may in reality be equivalent to muscles I and II 

 of C. bakeri. Its approach ventrally, in C. Jloridana, to touch muscle 

 II and then passing on into the peduncle may be somewhat similar 

 to the fact that muscle II in Cyclosalpa bakeri, if we correctly homo- 

 logize it, connects with both muscle I and muscle III. On the left 

 side of C. bakeri, muscles III and IV are united exactly as are the 

 muscles numbered II and III in Cyclosalpa Jloridana. All their 

 branchings are the same. This is true of the right side also, except 



