39 
the infero-posteal corner to an acute point;) inner ramus of up3 
about as long as 1. joint of the outer ramus. On the differences 
from G. locusta L. var. Zaddachi Sexton, vide infra, p. 41. 
Several specimens of this species have been taken, but only 
in the mouth of the fjord, 16—24 9% salinity, 0—10 m (a spec. 
from Mellerup, 6 miles inside the mouth, is too small to be deter- 
mined with certainty). The specimens correspond very well with 
Fig. 1. Gammarus loc usta Bu 16 mm, ns Udbyhøj. P5 > (prosimat part) and the 
third epimeral PR of the metasom 
Fig. 2. Gammarus locusta 9 with ov ma mm, from Mir tager Fjord. P4 and P5, 
proximal paris; the third iltre deg plate of the metasome, and the telson, 
Sars' description and figures; but as far can be seen all the speci- 
mens differ in having the infero-posteal corner of the 2. joint of 
PS much more rounded than shown by Sars, and not at all pro- 
duced, and almost rectangular; the form is in all my specimens 
from Randers fjord about the same as in G. pulex:" The 3. epi- 
meral plate of the metasome has spines on the under edge (not shown 
by Sars) and setæ on the hind edge; in the specimen from Ud- 
byhøj (fig. 1) (2, 16 mm; depth 8—10 m, salinity 24 "/oo) the 
hind corner is still more produced than in Sars' figure. 
In a spec. from Mariager fjord (fig. 2) (9 with ova, 18 mm; 
salinity 13.21 9/00) the third epimeral plate of the metasome has 
about the same form as in G. pulex, and the telson has almost 
totally lost all the spines and setæ: in the apex of the right half 
there are 2 spines, in the left half 1 seta. 
In order to investigate how much this species can vary I have 
revised some specimens of the material in the possession of our 
Zoological Museum (most of the spec. determined by Dr. H. J. 
