68 



Appendices to Second Annual Report 



(PI. IX. fig. 5), stones (PI. IX. figs. 1 and 2), or gravel (PI. IX. fig. 3), they 

 were always arranged in small, usually conical groups, which could only 

 have been formed by, the condensation of a continuous ribbon of eggs 

 deposited while the female was at rest. The lumps of spawn sometimes 

 found are probably formed out of the eggs which are thrown overboard 

 by the fishermen when cleaning their boats, or by the rolling action of 

 the waves during storms. 



In several instances, when a large dredge came up full of spawn, I 

 thought that I had at last come upon a part of the Ballantrae Bank, where 

 the eggs lay ' to a very great depth,' but on examination it was always 

 found that the spawn, instead of forming thick masses, was arranged in 

 irregular heaps ranging from a quarter to half an inch in thickness, — the 

 eggs nearest the gravel being often at a more advanced stage of develop- 

 ment than those on the surface, but owing to the openness of the arrange- 

 ment having a plentiful supply of water in contact with them (PI. IX. figs. 3 

 and 4). In such cases the sizes of the individual masses were sometimes 

 scarcely over an inch square, while at other times they were nearly 6 

 inches square. By laying the various portions side by side in a large tank, 

 it was possible to obtain a very accurate notion of the arrangement of the 

 eggs before they were disturbed by the dredge, and to feel satisfied that an 

 almost regular layer of eggs often cover several square yards of the bottom. 



In settling the question as to how the spawn is deposited, it is no longer 

 necessary to trust to the inferences made either from the examination of 

 the spawn dredged from the bottom, or from watching the somewhat un- 

 certain movements of the herring on the spawning ground, for I found it 

 possible to study the actual process of shedding the roe and the milt by 

 placing ripe herring from Ballantrae in tanks at the Bothesay Aquarium. 

 An account of the observations made is contained in the paper mentioned 

 above.* In this paper the natural spawning process is described as 

 follows : — 



Having secured at Ballantrae a large number of live herring, we f 

 selected the largest and ripest males and females, and placed them in a 

 large wooden tank, into which a number of stones and a quantity of sea- 

 weed had been previously introduced. After the fish had been about two 

 hours in this tank the stones and sea-weed were examined, and although 

 a few eggs were attached to them, it was quite evident that they 

 had not been deposited in the same way as those found on 

 the stones dredged on the previous day, but the presence of the eggs 

 indicated that we had secured ripe females. We were not surprised 

 that only a few isolated eggs were found on the stones, because the fish 

 had been disturbed every few minutes by the water poured into the tank 

 in order to keep the water sufficiently cool and abundantly aerated. 

 Having arranged a tent-like covering over the tank, so as to inclose the 

 fish in a dark chamber, we tried the effect of throwing instantaneous 

 flashes of light on the surface of the water, but as far as could be observed 

 this produced no impression whatever ; they neither sought the light nor 

 avoided it, neither did they seem in any way startled, however suddenly 

 it was directed across their path. This led us to believe that their 

 natural movements would not be seriously interfered with when intro- 

 duced into the tanks at the Bothesay Aquarium. 



On reaching Bothesay the hatching boxes and live herring were at 

 once transferred from the 'Jackal' to the tanks; a tank into which 



* ' On the Natural and Artificial Fertilisation of Herring Ova,' Proceedings Royal 

 Society, 1884. 



t Lieutenant Prickett, R.N., took part in this investigation. 



