REPORT ON THE BRACHYURA. 
107 
The genera comprised in the true Cancroid Cyclometopa are now extremely numerous, 
and, as is well known, are connected by almost insensible gradations of form and structure. 
Hence it is difficult, if not impossible, to indicate a system of classification whereby they 
may be divided into groups at once natural and distinctly defined; and the types 
represented in the Challenger collection are not sufficiently numerous to permit me to 
attempt (even did time and experience allow) a new and detailed arrangement of the 
genera of this group. 
The classification of Professor J. D. Dana , 1 although somewhat artificial, is one which 
is certainly very convenient to systematists, since this author gives a synoptical arrange- 
ment, with diagnostic characters, of all the genera known at the time when his work 
was published, and as it is the one which I have hitherto followed in my own papers, 
I have thought it best, pending a complete revision of this group, to adhere to this 
arrangement so far as the sequence of the leading genera is concerned in the present 
Report, separating merely the somewhat abnormal genus Trapezia (which approaches the 
Catometopa in the form of the carapace) from the remainder of the group ; but I refrain 
from indicating in the present imperfect state of our knowledge any subdivisional 
characters or subfamilies. 
A widely different and in some particulars more natural arrangement has been indicated 
by A. Milne Edwards , 2 who divides the true Cancridse into five primary, two transitional, 
and two accessory groups, characterised mainly by the form of the carapace, and typified 
respectively by the genera Cancer, Carpilius, Xantho, Eriphia and Trapezia, and by 
CEthra and Galene, and Pirimela and Liagore. His monograph of the recent genera of 
Cancridae has, however, unfortunately never been completed, and the limits and sub- 
divisional characters of six of the groups have not been defined, but it is probable that 
the true natural arrangement of the genera lies in the direction indicated by Milne 
Edwards. The genus CEthra (the type of his group CEthrides) is somewhat more nearly 
related to the Oxyrhyncha than to the Cyclometopa, and has been arranged, with Cryptopodia 
in the former group, both by Professor S. I. Smith and Dr. W. Stimpson, and by myself. 
Section I. Cancrinse. 
Canceriens arques, Milne Edwards, Hist. Nat. Crust., vol, i. pp. 369, 371, 1834. 
„ quadrilateres, Milne Edwards (pt.), tom. cit., pp. 369, 424, 1834. 
Carapace usually convex, with the antero-lateral margins arcuated, and armed with 
several lobes, teeth or spines. The front is of moderate width and usually does not 
project over the antennules and bases of the antennse, which are seldom excluded from 
the interior hiatus of the orbits. 
1 Crust, in U.S. Explor. Exped., vol. xiii. p. 145, 1852. 
2 Ann. d. Sci. Nat. (Zool.), ser. 4, vol. xviii. p. 39, 1862 ; Nouv. Archiv. Mus. Hist. Nat., vol. i. pp. 180, 182, 1865. 
