Contributions to the Crustacean Fauna of South Africa. 385 



contain a large amount of coagulated granular matter. At a later 

 stage, as seen in A. longicornis, this granular matter becomes greatly 

 reduced and is restricted to the basal and central portion of the plate, 

 which has become very much thinner and is surrounded by a perfectly 

 transparent border. In fact, in the 1st plate of A. longicornis by far 

 the greater portion of the plate is transparent. Consequently, I 

 believe, it is due to this transparency and the fact that the plate lies 

 flat against the ventral surface that this 1st pair of plates has been 

 overlooked. 



Possibly also the method of preservation may account for it. At 

 any rate, Koehler's statement must be regarded as incorrect. 



In Arcturella corniger (Stebb.) there are also definitely four pairs 

 of marsupial plates, as I have previously recorded in the description 

 of A. hirsutus, which proves to be a synonym of corniger (see below). 

 But in the species of Arcturella described below I believe there are 

 truly only three pairs of plates present.* 



On a further point also I am obliged to differ from the said author. 



The genus Arcturopsis was founded to receive certain forms which 

 were closely allied to Arcturella, but were said to differ in the presence 

 of a ventral process on the 3rd (or in one species on the 5th) peraeon 

 segment in the $ . This process was developed to a varying degree, 

 and was stated to occur in no other Astacillid (" il ue se rencontre 

 chez aucun autre Arcturide "). When describing Arcturopsis hirsutus 

 in 1914, I placed it in this genus on account of a small tubercle on 

 the 3rd segment in the <$ , which, though not developed to any such 

 size as in Koehler's species, was evidently homologous. 



On revising, however, all the species of Astacillids at my command 

 for the present paper, I found an exactly similar process on the 3rd 

 segment in Arcturella danmonensis (Stebb.) and in Astacilla longi- 

 cornis. I then applied to Prof. G. 0. Sars to know if such a process 

 was present in the type-species of the genus, Arcturella dilatata 

 Sars. Prof. Sars kindly examined his specimens and corroborated my 

 belief that A. dilatata did possess a process, albeit only " small, some- 

 what conical, anteriorly pointing" (Sars. in litt. 5/8/16). He also 

 confirmed my observation of its presence in A. longicornis. 



Dr. Caiman, to whom also I mentioned the matter, kindly examined 

 some species in the British Museum collection with the result that he 



* Note to p. 191 in Hansen, Dan. Ingolf. Exp., iii, 5, 1916. The presence 

 of a 5th pair of marsupial plates in Arcturus baffini is rather astonishing in 

 view of the strongly prehensile nature of the oth peraeopods and the position 

 assumed by the body. Hansen found 5 pairs also in two species of Pleuroprion ; 

 but this is less remarkable as this genus is more Idoteid in shape. 



