IN THE SOUTH-WESTERN DISTRICT. 143 



various phases of colouration, from a well-marked cupreous tint through 

 paler shades to a practical absence of any distinct colour at all. The 

 variation, in so far as I have observed it, usually affects individual 

 parents and not individual ova from the same parent, but, while pre- 

 paring these notes for press, I find among the ova just liberated by a 

 large Motclla imistcla some few with distinctly cupreous globules, while 

 those of the majority are colourless or only very faintly tinted. Its 

 explanation probably involves a physiological and chemical discussion, 

 which I am not qualified to enter upon. For my present purpose it 

 suffices to point out that the known existence of such a variation 

 renders it very unsafe to rely on resemblances or differences in coloura- 

 tion of the oil-globule for purposes of specific determination. I must 

 plead guilty to having done so myself, since the Motella Sp. iii. of my 

 Irish series {Trans. R Duh. Soe., S. II., iv., 1891, p. 464, PI. XLVII.; and 

 v., 1893, p. 95, PI. VI.) is chiefly based on the greenish colour of the oil- 

 globule. I must add that I have since found that this greenish colour is 

 identical with the olive-oil yellow of Piaffaele, the former being converted 

 into the latter by the use of a condenser. I am therefore of opinion 

 that my Species iii. can no longer be regarded as sufficiently 

 characterised. 



M'lutosh and Masterman {Life-Hist. Brit. Fish., 1897, p. 284) consider 

 that they can distinguish three species of rockling eggs in the tow-net 

 material of their district. I am in a less fortunate position here, for I 

 cannot find among the large number collected any distinctive character 

 which I consider absolutely reliable. Two rocklings, M. mustela and 

 M. tricirrata, are certainly common here, and no doubt the eggs of both 

 species have frequently come under the observation of Mr. Scott and 

 myself, but it has not so far been possible to check the tow-net material 

 by observation of the spawning of both species. I know that M, mustela 

 was spawning in March, 1897, while it has been taken full of roe in 

 January, 1898.* 



Of the spawning of M. tricirrata in this district I know nothing 

 definite; and the question is further complicated by the undoubted 

 existence in the district of M. clmhria and M. maculafa, and possibly of 

 other forms which may require specific distinction. I do not suppose 

 that the ova of M. maculata are small enough to be readily mistaken for 



* During the months mentioned females full of roe were seined in the estuary, at the 

 mouth of the Lynher in Jlarch, at the same place and also a little higher up the river in 

 January. A specimen transferred to the Laboratory, ou the 12th January, si>a\vued at least 

 as early as the 19th, since great numbers of eggs were found in the tank on that 

 date. On the same day rockling eggs were found in Plymouth Sound in water which Jlr. 

 Garstang pronounced to be estuarine in character, so that it is practically certain that 

 M. mustela spawns to some extent in the estuary. Rockling are known from Petersen's 

 observations to spawn in the Limfjord. 



