450 
ZOOLOGY OF THE FAR EAST. 
tions and figures are wanting in respect to the armature of the last three segments 
of the pleon. 
The only specimen of G. breweri as \'et collected had lost the third uropod and 
the published description gives no information as to the armature of the pleon. 
In G. piirpitrascens and G. propinquiis the inner branch of the third uropod is at 
least half as long as the outer and thus both species are very distinct from G. annan- 
dalei. 
I have not been able to consult the descriptions of G. polymorphus, Heifer, 1914, 
and G. argaeus, Vavra, 1906. With this brief review of the known species of the genus 
it is possible to state shortly the affinities of G. annaudalei. With the aid of Stebbing's 
key it is excluded under heading 22, and is thus related to G. pimgens, M.-Ed., G. 
ramelliis, Weckel and G. sowinskii, Behning. From these three species it is distin- 
guished by the shorter first and second antennae, by the form of the second and third 
thoracic limbs in both sexes and by the third uropods which have the inner ramus 
comparativelj'- longer than any of the above three species. It is, how-ever, as w^ell to 
point out that G. ramellus has the palm of the second and third thoracic limbs armed 
with the same type of peculiar spine as in G. annaudalei. But G. annandalei differs 
from all the species of the genus, in the possession of accessory vesicles on the bran- 
chial lamellae. Chilton (1916) in describing G. barringtonensis notes that " on some 
of the segments of the peraeon" there are finger-like appendages which appear to be 
of the same nature as the 'single accessory branchiae' described in Hyalella jelskii^ 
Wrzesn., and H. dybowskii, Wrzesn.", Chilton further says that these appendages 
appear to arise from the sternum of the segment internal to the branchiae, but he 
was unable to determine their exact occurrence. The processes seen by Chilton 
must be, I think, of the same nature as those I have noted above in Atyloides japonica 
and those seen by other aut:hors in species of Gammarus, Pontoporeia, and Synn- 
rella. They are quite distinct from the accessory branchial processes of G. annandalei 
w^hich are attached distinctly to the outside of the branchial lamellae. In no other 
species of Gammarus have I been able to find any mention of accessory branchial 
vesicles though they are found in the genus Hyalella and in some of the Eysianas- 
sidae. 
It is to be regretted that Smith's inadequate descriptions of G. ripensis and G. 
antipodeus do not permit of a closer comparison of these species with G. annandalei. 
They agree with the latter in the short inner ramus to the third uropods but Smith 
makes no mention of the armature of the pleon or of the structure of the branchial 
lamellae and it is not possible to say hoW' nearly allied to G. annandalei they really 
are. This is unfortunate because Smith regards these two species as in a measure 
intermediate in structure between the genera Gammarus and Neoniphargus and he 
suggests that the latter genus has been derived from the former in the Southern 
hemisphere and is not genetically related to the genus NipJtargus of the Northern 
hemisphere, the resemblance between Niphargus and Neoniphargus being regarded as a 
remarkable case of convergence. 
G. annandalei is a true Gammarus in all the characters that are supposed to dis- 
