320 



BULLETIN OF THE BUREAU OF FISHERIES. 



inside the second annulus and as the annuli seldom exceed two in number, branching 

 bears a much less important relation to age determination on the pigfish than on the 

 squeteague. At the focus the radii are usually from six to nine in number, and more 

 often than not continue to the periphery without branching. 



The radii on the pigfish scales corroborate the evidence found on the squeteague 

 and make even more convincing the probability that radii are adaptations to bodily 



movement. A cross section of the pigfish scale 

 shows a very characteristic structure strongly 

 supporting the proposed explanation. Imme- 

 diately under, and coincident with, the radii 

 are corrugations, in the upper layer of whose 

 summits are the radii. These corrugations 

 stain rose pink in picrocarmine (Hoyer), while 

 the superior layer stains hardly any. 



Scales taken from the opercula of the 

 pigfish offer still further corroboration of the 

 conclusions drawn from observations on the 

 scales of the squeteague. They present in 

 cross section a sinuous outline suggesting the 

 view that they are stages intermediate between 

 those that bend and have the regular radii and 

 those occupying inflexible parts and having no 

 radii. 



A series of experiments was carried on 

 directed toward determining what influence 

 food supply has on the formation of the annuli. 

 Two aquaria were kept. (See feeding record, 

 table 7.) In aquarium no. i four fish were 

 placed, ranging from 14 to 18 cm. in length; 

 in aquarium no. 2 seven fish of lengths ranging 

 from 14 to 22 cm. were placed. All were 

 started in aquaria on June 24. From June 27 

 the feeding record shows their treatment till 

 they died or were taken out, August 22. In 

 the beginning the fish in aquarium no. 2 were fed sparingly, but as it was learned that 

 they could live on very little, feeding was practically abandoned during the month of 

 August. In all cases the fish in aquarium no. 1 were fed daily all they would eat, while 

 those in no. 2, even when fed, were never satisfied. For some unaccountable reason all 

 of the "well-fed" fish died on August 6, while one of the "starving" fish died August 5. 

 Careful watch was kept on the growth of the scales, but no difference whatever 

 was noticeable in the formation of annuli. Any difference in growth was so small as 

 not to be reckoned with. (PI. Lviii, fig. 25, 26.) 



These results appear to offer conclusive evidence that feeding habits have no in- 

 fluence upon the formation of annuli. Other possible factors yet untested are the 

 influence of temperature and the presence of lime salts in the food and water. 



2 3 



Age in years 



Fio. 6. — Occurrence of pigfish at different ages. 



