St. Andrews Marine Laboratory. 101 



are not uncommon amongst the specimens at St. Andrews — 

 some of these being only four inclies in length, or even 

 somewhat less. 



The ova of this species agree in general structure with 

 those oi A. lanceolatus. In the early condition the capsule 

 {zona radiata) encloses only minutely granular yolk, no 

 special oil-globules being visible. The latter subsequently 

 become distinct, and by-and-by coalesce into a single large 

 oil-globule, as in the latter species. The colour of the 

 globule, however, is different, being of a honey-colour or 

 faint yellowish brown. 



The spermatozoa are more minute than in A. lanceolatus^ 

 but appear to have the same shape, viz. like a slightly curved 

 grain of rice, no perceptible diminution occurring at the end 

 from which tlie filament proceeds. Their vitality would 

 seem to be as feeble as in that species. 



Artificial fertilization was frequently performed with speci- 

 mens sent by the courtesy of the Fishery Board from Elie, 

 but at first without success. It is true certain changes 

 ensued, but whether these were altogether due to partial 

 fertilization, is an open question. A slight streaming of 

 protoplasm occurred with the formation of a thick belt round 

 the yolk. The perivitelline space also largely increased in 

 size. Then a conical elevation of the protoplasm formed a 

 disk, but no segmentation took place. The persistent feeble- 

 ness of the sperms after the journey showed that it was 

 necessary to fertilize on the spot — immediately after landing 

 at Elie. This was accordingly done, with the result that 

 after the usual streaming of tlie protoplasm on the surface of 

 the yolk, the increase of the perivitelline space, and the for- 

 mation of the prominent conical disk, segmentation ensued. 

 Towards the end of July a considerable number of ova were 

 successfully hatched, so that the larva was satisfactorily 

 identified with those abundantly procured by the various nets 

 from the early part of the year onwards. l)r. Fullerton also 

 hatched them at the Laboratory at Dunbar at the same 

 time. 



The constant employment of the various tow-nets at dif- 

 ferent depths throughout the year shows that few forms of 

 larval, post-larval^ and young fishes are more abundant, 

 more generally distributed, or occur over a longer period than 

 the sand-eels. At St. Andrews they appear in February, 

 and the larval and post-larval forms are found onward 

 through March, April, May, June, July, and August, show- 

 ing that a constant succession of eggs and young are kept up 

 throughout this long period. 



