24 /. H. Powers 



feeding in water. Further, it is interesting that this breadth of 

 tail was not due, as I have said in speaking of the larva, to the 

 swimming habit, but to the very opposite cause, viz., to rapid de- 

 velopment at the bottom, with excessive nutrition. The slender- 

 tailed adult is the result of slower growth, mostly in a cistern, 

 where the specimen metamorphosed after about a year and a half 

 of slow growth. It is therefore an example of form produced by 

 slow growth, and to some extent of the free-swimming habit. 

 The larva from which this specimen resulted was similar to figure 

 4 on plate 11. This specimen was not photographed until several 

 months after metamorphosis, and in order to test the permanence 

 of these slender-tailed adults, this individual and several like it 

 were kept in water and fed to the utmost, alternately on meat and 

 on daphnids, for an entire season. The specimens became very 

 fat, to judge by external appearance, but none underwent sig- 

 nificant tail expansion, this being one out of many instances in 

 which I have observed the tendency of rather extreme forms to 

 remain permanent, or even to increase their variation rather than 

 develop toward a supposed norm. 



I may now add a word of explanation concerning the broad 

 tails of figures 2 and 3 on plate V. Figure 3 represents a rather 

 slender adult, not unlike figure i, plate ITI, at the time of its 

 metamorphosis. After being kept in water for nearh' a year sub- 

 sequent to metamorphosis, the tail underwent, rather suddenly, 

 this expansion, the impulse being plainly due to a period of heavy 

 feeding. The special form of this tail is rather unlike most of 

 those among my water specimens. Figure 2 is unfortunately pho- 

 tographed from a preserved specimen which has shrunken some- 

 what. But it shows my most aquatic adult, after about two years 

 of life in a large battery jar where it had several companions. It 

 will be noticed that the tail is very long, somewhat more than 

 equaling the length of head and body. It is also very high and 

 thin, though scarcely finned. In life the extremity of the tail, 

 aside from the notochordal axis, was as thin as a leaf and the tip 

 was pointed like that of the average larva. Some larvae even 

 have rounded tail tips, like figure 2 on plate I, while most end in 

 a pointed tip like figure 3 on plate I. The shape of the tail in this 



220 



