1897 - 98 .] Prof. D’Arcy W. Thompson on Marine Faunas. 347 
1 Challenger ’ Expedition is scarcely identical with Boeck’s 
species.” 
The difficulty and uncertainty attending the examination of 
forms, even so highly organised as these, may he illustrated by 
the following instance. 
The species now called Polycheria antarctica , originally described 
by Stebbing from a dried specimen taken from a sponge in the 
Antarctic Ocean under the name Dexamine antarctica {A. M. N. 
H., (4) xv., p. 184), later transferred by him to the genus Atylus 
{A. M. N. H., (5) ii., p. 370), and now regarded by him as 
identical with the form which he afterwards described ( Chall . 
Rep., Amphipods) from Kerguelen as Tritceta kergueleni (under 
which name it is cited in Murray’s recent list as one of the 
species peculiar to Kerguelen), is in all probability identical with 
the forms described from Australia and New Zealand under the 
names Polycheria tenuipes , Haswell, P. brevicornis , Haswell, and 
P. obtusa, Thomson. In other words, this single species has been 
described by so great and careful an authority as Mr Stebbing as 
two species under three generic names, now replaced by a fourth, 
and by other writers as three other distinct species. 
Of the genus to which this species is now ascribed, namely, 
Polycheria, , one other species only is known, from Puget Sound, in 
Western Korth America. 
SUMMAKY. 
The foregoing analysis shows that of some ninety species quoted 
by Murray as common to northern and southern localities while 
absent from the intermediate zone, there are more than one-third 
in which grave doubt as to their identification was expressed by 
the original describers, or in which the identification has been 
doubted or denied by later writers. 
In somewhat more than another third the evidence of identity 
is inconclusive or even inadmissible by reason of the nature of 
the examination to which the specimens were subjected (as in the 
case of the horny and calcareous sponges), or by reason of the 
small size of the objects and lack of adequate marks of charac- 
