144 NOTES ON THE RErRODUCTION OF TELEOSTEAN FISHES 



those of other British species, but M. cimhria does not seem to be a very- 

 large form. The pelagic larval rockling, commonly known as " Mackerel 

 midges," would afford more assistance to a knowledge of the spawning 

 season if it were possible to identify them with absolute certainty, but 

 it seems quite possible that those usually associated with 31. tricirrata 

 may not all belong to that species. 



On the whole I do not think it would be profitable to enter at present 

 upon a detailed discussion of the probable parentage to the tow-net ova. 

 It may be noted that the eggs with colourless globules correspond in 

 character with the descriptions of M. m.usida, and in date with the 

 known spawning period of that species at Plymouth. The cupreous 

 colour of the globule, noted in several ova by Mr. Scott in January, is 

 the same as has been referred to above as observed in newly extruded 

 ova of the same species. The yellow, which under different conditions 

 of illumination is either the " olive-oil " of Raffaele or the green of my 

 Irish notes, has not been noted in ova directly derived from the female 

 of any species. The ova with yellow globules, first observed in January, 

 continued to occur until the middle of September, and after April were 

 much commoner than those with colourless globules, of which the last 

 was observed in June. If the yellow colour is really a constant character 

 and occurs only in one species, then that species must have a spawning 

 season of nine months. I have failed to recognise any differences in the 

 pigmentation of larvae with colourless and yellow globules respectively. 

 In our records the colour of the oil-globule is stated whenever it was 

 noted. The record of dimensions was very insufficiently kept by myself 

 during the later months of the year, and it is partly on this account 

 that I defer a discussion of this part of the question. It appears, 

 however, sufficiently plain that there is considerable variation, not only 

 in the diameter of the egg, but in that of the oil-globule, proportionally 

 as well as actual, in ova which are similar in colouration of the globule 

 and in embryonic and larval characters. . 



Mr. Scott's notes contain references to, and a drawing of, an egg with 

 very numerous oil-globules. It is identical with a form described by 

 M'Intosh and Masterman (p. 396, PI. IV., Fig. 13) as closely allied to 

 Solca lutea, and is in reality the egg of a rockling, probably liberated 

 before it was perfectly ripe. I have seen very similar ova at St. 

 Andrews, which ultimately, by coalescence of the oil-globules, assumed 

 the ordinary appearance, and have also obtained them at Plymouth 

 directly from a female of M. mustela. I do not think that any of the 

 tow-net specimens were fertilised. 



A " definite pale area, slightly refractive and apparently differentiated 

 from the yolk," noticed by M'Intosh and Masterman (p. 296) in the egg 

 of a rockliug, is a common feature in unfertilised eggs of G. luscus. It 



