362 REPORT OF COMMISSIONER OP FISH AND FISHERIES. 



ment no longer than the others as noticed by Johnston*, and later 

 by Stebbingt, who draws from the fact an argument against the validity 

 of the genus. I fail to see, however, why the argument would not be 

 equally valid agaiust the use, among mammals, of characters drawn from 

 the horns and teeth. Nothing is more common, in case of a genus or 

 family possessing a special development of some organ or set of organs, 

 than to find that the young of sucli a group resemble the adults of less 

 specialized groups. If, however, as may be possible, a gradation can 

 be established between forms which, like Arcturus Baffini^ have the fourth 

 thoracic segment large but only slightly elongated, and forms like Asta- 

 cilla longicornis or A. granulata, in which this segment is much elongated, 

 equaling or suri)assing the other six in length, there would then be, per- 

 haps, no sufficient reason for retaining both genera. For the present it 

 seems desirable to keep them separate, and to the characters given above 

 we may add the following : 



The head is produced at the sides around the bases of the antennulre, 

 and is united dorsally with the first thoracic segment, the sutures being 

 evident only at the sides where the segment is produced around the 

 hinder part of the head. The flagellum of the antennse consists of three, 

 or sometimes only two, distinct segments and a terminal spine, which is 

 perhaps to be regarded as a third or fourth segment. The maxillipeds 

 (pi. IX, fig. 52 a) are robust and opercuUform, with a thick external 

 lamella and a five-jointed palpus, but little flattened. The mandibles 

 are destitute of palpi. 



The first three thoracic segments are subequal and short ; the fourth 

 much elongated in both sexes ; in the males it is slender and cylindrical ; 

 in the females it is more robust, and bears on its inferior surface the incu- 

 batory pouch. This pouch is thus confined to a single segment, and is 

 composed of a pair of elongated lamellce, attached along their outer mar- 

 gins, and overlapping widely along the ventral surface. It occuijies 

 nearly the entire inferior surface of the segment. The last three thoracic 

 segments are short aud subequal, and the articulation at the posterior 

 end of the fourth segment is capable of considerable motion, and, in our 

 species, is usually flexed backward nearly at a right angle. The first pair 

 of legs (pi. VIII, fig. 49 h) have the basis directed backward and the re- 

 maining segments ciliated and turned forward, and is more robust than 

 the three succeeding pairs, which are slender, of nearly equal size, and con- 

 sist of only five segments, which are turned forward from the basis and 

 held beneath the head. They are strongly ciliated, especially on the last 

 three segments. One of the fourth pair of legs is shown on plate VIII, 

 figure 50. The last three pairs of legs are of entirely different structure, 

 being robust and prehensile with strong short dactyli. 



The pleon is consolidated into a single segment, which, however, shows 

 traces of its composite nature. It is vaulted above and excavated on 



*Loud. Mag. Nat. Hist., vol. ix, p. 81, fig. 15, 1836. 

 t Ann. Mag. Nat. Hist., IV, vol. xv, p. 187, 1875., 



