1878.] MR. R. COLLETT ON CERTAIN GOBIOID FISHES. 329 



strings take up the greater part of the belly, without, however, as is 

 the case in Crystallogobius nilssonii, reaching behind the vent along 

 the root of the caudal fin. The number of the eggs is between 1 800 

 and 2000. 



Habits. — This species, as well as the following differs in its habits 

 in several respects from most of the typical Gobiidte. Above the 

 compressed body the laterally placed eyes indicate that they do not 

 live at the bottom, as is the case with the greater number of 

 Gobies, but swim about freely. Neither do they keep themselves 

 dispersed or in single numbers, but in large and dense shoals, 

 where their perfect transparency, in connexion with their inferior 

 size, enables them to avoid immediate observation. Sometimes I have 

 taken them from the stomachs of other fishes ; but by far the 

 greatest number are brought up to the surface by the nets set out 

 for different other fishes, especially for the sprat and the young of 

 the common herring. These fishes are caught in the Christiania 

 Fjord in September and so long as the fjord is free from ice — as a 

 rule, until the beginning of the new year. In the spring and summer, 

 when the nets are drawn for mackerel, they are also to be found, 

 and then with the organs of generation fully developed. These nets 

 are drawn at a depth varying from one to fifteen fathoms. An 

 examination of the stomachs of these fishes shows that they contain 

 chiefly pelagic copepods and the fry of mollusks in their swimming- 

 stages ; these animals principally live in the upper strata of the sea 

 some few fathoms under the surface. The bottom consists, in most 

 of the inner parts of the Christiauia Fjord, of clay and mud, which, 

 in the more shallow water, is covered with Zostera. 



These fishes seem always to keep together in enormously large 

 shoals ; and when the meshes are very fine they are sometimes caught 

 in great masses. The 30th of October 1875 (which was a par- 

 ticularly successful year for their development), I found in a single 

 draught (which, besides, also brought in some young specimens of 

 two or three species of Gadus, several Pleuronectes Jlesus, and Cteno- 

 labrus rupestris, some Gobii of common species, and herrings) such 

 large masses of them that I collected between two and three thousand 

 specimens, whilst possibly more than double this number had already 

 escaped through the meshes. In the same year I was not able to 

 notice any diminution of their numbers, and they were always found 

 in the same places ; still they are not at all equally numerous every 

 year, and they have been rather scarce both in 1876 and 1877, es- 

 pecially in the latter. 



In the Mediterranean they must be produced in still greater 

 numbers, as they are caught in very fine nets and sold as food. 

 In several of the fishing-markets in the towns by the Adriatic and in 

 Sicily Prof. G. O. Sars saw vessels filled to the brim with these 

 fishes alone, forming a complete jelly-like mass. Kessler also men- 

 tions their use as food in Odessa. 



Upon the least touch the scales fall off; and after the nets are 

 drawn, scarcely a single specimen will have its scales complete, and the 

 greater number will have lost almost every scale. The proportionally 



