192 BULLETIN OF THE 



as shown in Figs. 11, 12, nnd 13, has become reduced in size, so thai now tlie 

 greatest diameter is Ibund just l)eliind the cephalic sacs, and the body tapers 

 from this point backward to its posterior end. The hirva has three, pigmented 

 regions, viz. the very anterior end of the snout, the region just behind the 

 cephalic sacs, and the posterior end of the body. The eye-spots are found on 

 the sides of the head just in advance of the cephalic sacs. These last-men- 

 tioned organs are relatively smaller than formerly, and are ciliated in their 

 interior and on the inner walls of the tubes which lead to them from the 

 external orifices. 



The internal organs differ slightly from those of the adult Lincus. A large 

 stomach, the movements in the walls of which could be easily seen, fdls most 

 of the interior of the larva. There is in this larva no visible anus. The sinu- 

 ous tube (ic () which lies on the dorsal side of the stomach has been identified 

 as a water vessel. A similar organ is figured by Leuchart and Pagenstecher in 

 P. (jtjrans. 



The form of this singular larva and the strange development of the enclosed 

 worm suggest very interesting theoretical questions. The external outlines of 

 the larva ally it to animals widely separated in our classification from the 

 Kemertean to which it gives origin. Busch, Leuckart and Pagenstecher, and 

 others, have pointed out that in the similar relationship of Ncmcrtes to its 

 Pilidium we have a parallel condition to that which exists between an Echinus 

 and its fluteiis. In the species of Pilidium which they studied, not only the 

 stomacli of the larva, but also its oesophagus and mouth, were found to be 

 directly changed into the same organs in the contained Nemertean. In P. re- 

 curvatum the resemblance which they suggested is even closer than in those 

 species which they have studied, for here not only is the larval mouth and 

 (Esophagus only indirectly if at all changed into the same organ in the adult, 

 but also the lower portion of the embryo has a true brachiolarian form which 

 is highly suggestive. Our larva, even more than those of other species of Pili- 

 dium, shares with Tornaria and Adinotrocha many Echinoderm characteristics. 



The history of the opinions which have been advanced by Miiller, Busch, 

 Leuckart and Pagenstecher, Huxley, A. Agassiz, and others, in relation to 

 the resemblance or want of likeness of Tornaria to the young Echinoderms, 

 is too well known to be repeated here. If Balanor/lossus were the only worm 

 whose larva resembles the young Holothurian, the fact might be explained by 

 the abnormal character of the adult. With the Xemerteans, however, the case 

 is somewhat diff'erent, for in them we have a large group, whose larvae have 

 many points of resemblance to the embryonic Echinoderm. Nowhere is that 

 likeness carried so far as in the strange Pilidium, recurvatum, which has been 

 described above. It would seem at first sight that the circular belt of cilia 

 described in this larva would be an argument against its close affinity with the 

 larval Echinoderm. The same thing may be said of this, which has already 

 been said of a similar belt in Tornaria. Far from being an unknown feature, 

 it is a peculiarity in some Echinoderm young, as in Comatula and the Holo- 

 thurians, which are the closest allies of the worms. We find in some young 



