and Mysidellinte. Three of these subfamilies, viz. the two first-named and the Mysidellins, are 

 accepted here, while the four others, separated by too vague or unimportant characters and 

 scarcely maintainable, are put together constituting the subfamily Mysina;, which again is sub- 

 divided into tribes of lower rank. In 1905 Holt & Tattersall (Schizopodous Crustacea from 

 the North-East Atlantic Slope: Fisheries, Ireland, Sci. Invest. 1902—3, IV) established the 

 subfamilies Arachnomysinse and Boreomysinae, the latter of which I accept. In 1906 the same 

 authors in a supplement to the paper named (Fisheries, Ireland, Sci. Invest. 1904, V) established 

 the subfamily Mysidetinse which I cannot accept, and in 1909 Tattersall (Mitth. Zool. Stat. 

 Neapel, 19. B.) established the subfamily Calyptomminae on immature specimens of a single 

 species, wherefore a final judgment on this group cannot be given. 



Having examined representatives of the majority of the numerous genera of the family 

 Mysidffi hitherto known I think it useful to attempt to arrange them as naturally as possible. 

 It may be convenient to begin with a brief examination of all parts of the skeleton in order 

 to be able to estimate the systematic value of the differences. Then it will be more easy to 

 point out some main lines for the classification as the base for the establishment of subfamilies 

 and the subdivision of one of these into tribes. It is, I think, generally admitted, that the sub- 

 order Lophogastrida is lower than the suborder Mysida ; the latter is very naturally divided 

 into the two families : Petalophthalmidse and Mysidse, and the former family is lower than the 

 latter, showing some characters rarely or not at all met with in the Mysidse but found in the 

 Lophogastrida, as the number of marsupial lamellee, the undivided state of sixth joint of the 

 thoracic legs, no otocysts, etc. Therefore the Lophogastrida and the Petalophthalmidae are of 

 importance to the correct estimate of several structural features in the Mysidse. 



The Cephalothorax is essentially uniform in most genera of the family. In Arach- 

 nomysis Chun the first thoracic segment is extremely elongated, with the result that there is 

 a very long distance between the maxillipeds and the gnathopods. In LongitJwrax Illig and 

 especially in Gymnerythrops n. gen. the last segment is extremely elongated. But an investigation 

 of the structure of these genera shows that the cephalothoracic characters are only secondary 

 modifications. 



The Carapace shows considerable variation. In some forms (y^wf/zza/wm Norm.) it covers 

 the posterior cephalothoracic segments, but in most species it is considerably emarginate behind, 

 leaving uncovered the major dorsal part of the last segment and frequently besides of the 

 penultimate segment. Anteriorly it is most frequently produced into a projecting, less or more 

 rounded or acuminated, triangular plate which in Siriella frontalis j\I. Edw. and S. armata 

 M. Edw. is somewhat long, but sometimes the front margin is only a little convex or feebly 

 angular ; in Arachnomysis Chun it is anteriorly shorter than in the other genera, with the front 

 margin transverve and armed with spiniform processes. In most species of Borconiysis G. O. S. 

 the lateral part of the front margin is produced below into a conspicuous lateral plate. In 

 Arachno7nysis Chun, Chnnomysis Holt. & Tatt. and especially in Ccrsaroinysis Ortm. the carapace 

 has a number of spiniform processes. The particulars mentioned are frequently valuable as 

 generic or specific characters, but scarcely any one among them may be considered of higher rank. 



The Eyes with their Stalks present many differences, but all at most only of 



