I lO 



not yet twice as far distant from the 2"*^ as the 2^^ from the 3'^'^. In this species the fingers 

 of the chelate legs are a trifle shorter than in P. japoiiicus. 



The rostrum of the female from Stat. 213, that is 78 mm. long, reaches to the end of 

 the penultimate joint of the antennular peduncle and is 7-dentate; the 5"^ tooth is placed above 

 the orbital margin, the i*' is twice as far distant from the 2°^^ as the 2°'^ from the 3''^. The 

 female from Stat. 234 has nearly the same size, the tip of the rostrum is broken off. 



The female from Tiur-island measures 67 mm., at this age the two lobes of the thelycum 

 are not yet in contact, but separated by a narrow fissure from one another. 



Remarks. The type of P. plebejtis Hess from Sydney is a well-preserved female, 

 nearly 15 cm. long. P. catialiculatus var. australiensis Sp. Bate, the type specimen of which 

 I have observed during my visit to the British Museum in the summer of 1908, is identical 

 with this species. Like P. japojiicus, also P. plcbejtts Hess ought to be considered as a proper 

 species, not as a variety of P . canaliculatjis. The rostrum, 7-dentate, reaches almost to the 

 end of the 3"^ antennular article. The tooth on the lower margin is situated just below the 

 foremost tooth of the upper. The groove of the rostral carina and the two lateral grooves 

 resemble those of P. latisitlcatiis, but one observes, at either side of the rostrum, about midway 

 between the lateral rostral carinae and the tips of the teeth, another carina that commences 

 near the base of the first rostral tooth and that, gradually narrowing, is continued to the 

 foremost tooth, where it passes into the upper margin of the rostrum. This second carina is 

 wanting in P. latisulcatus. 



In P . canaliculatus^ P. japoniciis^ P. latisitlcahts and in P. plebejus there is a small 

 post-ocular spine, continued as a ridge that runs parallel with the rostral crest ; posteriorly, as 

 Dr. Alcock describes for P. Japojiic7is (I.e. 1906, p. 15), this ridge is, in all these species, 

 recurved on itself to form a narrow loop. It is also observed in P. carainotc of the Mediter- 

 ranean, but in P . brasilie7isis Latr. such a loop does not e.xist. In the type of P. plebejus, 

 however, there are two loops, instead of one, between the posterior part of the post-ocular 

 ridge and the rostral crest ; of the two loops that are separated from one another, from the 

 rostral crest and from the post-ocular ridge, by deep, though narrow, sulci, the inner is a little 

 shorter than the outer. As regards the other grooves and crests on the carapace P. plebejtis 

 accords with P. latisulcahts. In P. plebejus the carpal joints of the thoracic legs show a 

 somewhat stouter shape than in P. latisulcatus, but the difference is slight and only 

 recognizable by accurately comparing the two species. 



The thelycum, finally, has a somewhat different form from that of P . latisulcatus. The 

 two lateral plates or lobes come in contact with each other at the median line, as in P. lati- 

 sulcatus, the distal third or fourth part excepted, but their form is semi-elliptical, and their 

 lateral margins, that are slightly curved, run nearly parallel, whereas in P. latisulcatus they 

 distinctly diverge backward. The lobes show about the form figured b)- Bate (Challenger 

 Macrura, PI. XXXII, Fig. 3"), except that in this figure the lobes are not in contact. In P. 

 latisulcatus the two lobes are therefore broader at their base than in the middle, whereas in 

 the australian species they are a little broader in the middle than posteriorly. The median 

 protuberance differs likewise. Both in P. latisulcatus and in P. plebejus it consists of a posterior 



