4S1 



specimens at Haugesund. It is by no means very active in its movements, 

 and, when swimming about, carries the 2 anterior pairs of pereiopoda spread 

 oat in a peculiar manner to each side. 



Distribution. Shetland Isles (Norman), British Isles (Robertson), 

 coast of France (Chevrenx), Adriatic (Grube), Mediterranean at Messina (the 

 author), Azores (Barrois). 



Fam. 20. Gammaridae. 



Body more or less slender, with the segments of urosome well denned. 

 Coxal plates of moderate size, or very small. Antennas generally rather 

 slender, and, as a rule, but little different in the two sexes, the superior 

 ones provided with an accessory appendage often greatly developed. Oral 

 parts normal, except in the genus Lilljeborgia. Gnathopoda generally rather 

 powerful and subcheliform, being, as a rule, much more strongly built in 

 male than in female. Pereiopoda more or less slender, the 3 posterior pairs 

 generally increasing in length, and having the basal joint laminar. Last 

 pair of uropoda, as a rule, projecting beyond the others, and having 

 the rami more or less foliaceous. Telson generally small, lamellar, and 

 more or less deeply cleft. 



Remarks. The present family comprises a great number of Amphi- 

 poda, which on the whole may be said to exhibit most prominently the 

 typical features of the order. Yet there are some forms, which differ more or 

 less from the type, as revealed in the genus Grammar us, and which exhibit, 

 as it were, transitions to other families. Thus the genus Mclpliidippa would 

 seem to approach the Atylidfr, whereas the genus Lilljeborgia shows some 

 resemblance, especially in the structure of the oral parts, to the Leucoihoida, 

 with which family it was in fact associated by Boeck. Perhaps the 2 above 

 named aberrant genera should more properly be wholly removed from the 

 present family, and be regarded as types of 2 separate families. In such case 

 indeed, the family Gammaridce should have been much more precisely denned. 

 I feel, however, some hesitation in increasing the number of families, especially 

 where such families would be founded only upon isolated genera, and I 

 therefore prefer provisionally to include the said genera in the family Gain- 

 maridcB, placing the one at the head, the other at the end. The family com- 

 prises 12 Norwegian genera, two of which are now for the first time 

 established. 



63 Crustacea. 



