696 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii. 



cately and fully branched, having a very bushj^ appearance. It is in 

 relation with an extensive and conspicuous rete mirabile of the intes- 

 tine. This rete mirabile, which is as well developed as in some 

 species of Holothuria., begins about 30 mm. behind the diverticulum 

 of intestine and extends over 100 mm., occupying nearlj- all of the 

 first large, backwardly directed U-shaped tract of the intestine. The 

 vessels of the plexus are wonderfuU}' numerous, being larger in 

 the anterior portion of the rete mirabile than in the posterior (with 

 reference to intestine, since the morphological posterior end of plexus 

 is a little anterior to the front end, on account of U-shaped twist 

 of alimentary canal). Longitudinal muscle bands divided, the mid- 

 ventral strand the narrowest, the two dorsal remarkably wide, 

 about twice as wide as the mid- ventral. The ventrolaterals are inter- 

 mediate in size. Cloacal dilation is large, extending about 40 mm. 

 forward from anus. 



This species differs from true Paelojpatides in two very important 

 particulars, namely, in the possession of ventrolateral pedicels in the 

 hinder half of body (these being absent in typical P^lopatides) and in 

 having a well-developed rete mirabile. One of the characters given 

 for the subfamily Synallactinje is the absence of a rete mirabile. 

 There can be no doubt, however, that the present form is a Pselo- 

 patides or nearly related genus, because the outward habit, lack of 

 free tentacle ampulla^, two unequal Polian vesicles, large intestinal 

 coecum, and absence of calcareous deposits all point to it. 1 am not 

 so sure that the type of the genus is without a rete mirabile, for Theel 

 says" in the description of confundens: "The left [respiratory] tree 

 is shorter but more branched, its ramifications being in communication 

 with the plexus of pseudh^mal vessels. The right tree attains almost 

 the length of the body itself." So far as the trees are concerned 

 this agrees exactly with retifer. Ludwig makes no comment on this 

 in his diagnosis of the Synallactinte. His specimens of confundens 

 seem to have lost the left tree and most of intestine, so that naturally 

 the rete mirabile would be missing. Sluiter, who lists P. confundens 

 in his Siboga Holothurioidea, does not mention the rete mirabile or 

 make any reference to the apparently unnoticed remark of Theel 

 above quoted. Inasmuch as Sluiter's specimens lacked the intestinal 

 diverticulum I think there are grave doubts as to their being true 

 confundens. Theel's t3^pe came from off the coast of Chile, in the 

 latitude of Valparaiso, while Sluiter's specimens were taken in the 

 distant East Indian region (between 116° and 132° E., and 0°-8° S.). 

 Retfer appears to be distantlv related to P. purpnreo-punctidiis 

 Sluiter. It differs from this form in having a single continuous 

 series of papillge all along the ventrolateral radii, in addition to about 

 ten large pedicels, which form a separate series parallel with the above 



« Challenger Holothurioidea, Pt. 2, p. 156. 



