of the Fishery Board for Scotland. 



217 



any probability in favour of a decrease or an increase in length. 

 Maturity also lias no effect on this character. Hence, we may conclude 

 that the body-height and the length of the tail do not alter relative to 

 the body-length during growth. 



Skull-length. — This dimension, as Table XIII., p. 232, demonstrates, 

 shows a decrease due to growth-variability in all the groups. If the 

 groups of mature specimens alone be considered, then a comparison of 

 three groups between 250 -350mm. with other three between 350mm. 

 and upwards is possible. In the first two groups — those from the 

 southerly North Sea — the probability is about 1000 to 1 in favour of a 

 decrease in the relative length of the skull. In the case of the third — 

 the male specimens from Aberdeen— the probability in favour of this 

 decrease is only about 4 to 1, but it will be noticed that the number of 

 specimens is small in comparison with those of the former groups, and 

 this, as already mentioned, lowers considerably the amount of pro- 

 bability. 



The decrease is for the three cases '88, 1*01, and '36. The average 

 of these is *75, so that there is a decrease of "75 per cent, in the 

 length of the skull relative to the length of the body in mature speci- 

 mens. The difference in the total length of the fish is approximately 

 100mm., because the average for the specimens between 250- 350mm. 

 is 300mm. in both cases, whilst that for those over 350mm. is just a 

 little under 400mm. This decrease in the skull-length, it will be noticed, 

 applies both to males and females. 



If the first two columns be now considered, those containing the 

 immature and mature specimens of the same size between 250-350mm., 

 it will at once be noticed that the differences in the averages between the 

 two groups are smaller than were those between the mature specimens 

 at different sizes. The average difference, indeed, only amounts to '42 

 per cent., where above it was '75. Again, if the probability of the 

 observed decrease be calculated, it is found to be 19 to 1 in the first 

 two cases — i.e., those from the southerly North Sea, but only 1 to 1 in 

 the third case — i.e., the male specimens from Aberdeen. Hence, the 

 probability that maturity, or the ripening of the reproductive organs, 

 affects the length of the skull is comparatively small, and the actual 

 decrease observed is small in comparison with that observed as due to 

 growth- variability. 



Both Matthews and Heincke found that the length of the skull 

 relative to the length of the body decreased in the herring during 

 growth, but do not give the amount of decrease. The comparison 

 between the herring and the plaice — a free-swimming round fish with a 

 semi-sedentary flat fish — would, on this point, as on many others, be of 

 considerable interest. Such a comparison will be possible later when the 

 promised second portion of Heincke's work appears. 



Shut I -breadth. — This dimension was measured from the pterotic of the 

 one side to the pterotic of the other, near the posterior extremity. In 

 the plaice the posterior end of the pterotic slopes backwards and inwards 

 to form an articulating surface for the second head of the post-temporal 

 bone of the pectoral arch, and thus presents no definite end-points for 

 measurement. Immediately in front of this articulating surface is the 

 last tubercle on the pterotic, and as this tubercle offers a definite end- 

 point the dimension was measured from tubercle to tubercle. The varia- 

 tions in height of these tubercles are thus included under the variations of 

 the skull-breadth. 



As with the skull-length, Table XIV., p. 233, shows that there is a 

 decrease in the breadth of skull relative to the increasing length of 

 body, The proportions given in the Table are relative to the skull- 



