44 /. H. Powers 



not from varying degrees of the swimming habit, but from var3ang 

 degrees of nutrition. The broad foot, with broad depressed toes, 

 oval or even triangular in outline, is a product of what I should 

 call overgrowth, of excessive anabolism; whereas the longer and 

 much more slender toe is the well-nigh invariable result of pro- 

 longed, slow, or moderate growth. Compare figures 8 and 9, on 

 plate II. The former shows the hind legs of a robust twenty- 

 centimeter larva out of the same lot as are figure i,- plate I, and 

 figure 2, plate II. Its life had been passed as a bottom crawler, 

 not as a swimmer, and it had attained its massive size by the 

 middle of August, the eggs having, in this case, probably been 

 deposited in April. Figure 9 shows the posterior limb and foot 

 of a slender larva, 27 cm. long, which had grown slowly and con- 

 tinuously for about eighteen months. Such examples are thor- 

 oughly characteristic. Figure 7, plate II, shows another slender- 

 toed foot of a slender, slowly grown larva which I had raised. 

 Figure 6 shows a characteristic foot of a slender pond larva of 

 slower growth, the larva being but a trifle shorter than that from 

 which figure 8 was taken and showing an equal preparation for 

 metamorphosis. Figure 10 shows the posterior foot of a mam- 

 moth sexually mature larva, one year old, and 27 cm. in length. 

 It had outgrown all its companions by a centimeter or more in 

 length, this superiority being, however, much less relatively than 

 its excess of growth in peripheral parts, viz., in parotid glands, 

 gills, and breadth of toe. 



But are these depressed toes with bordering skin folds retained 

 by the adult? Weismann includes the cutaneous margins among 

 the larval organs which are lost during m.etamorphosis. Obser- 

 vation of many instances shows that this may or may not be the 

 case : frequently there is a little absorption, less frequently there 

 is a good deal, or none at all. Moreover, in cases of this char- 

 acter, metamorphosis accentuates rather than reduces the larval 

 variation, but, curiously enough, in a manner exactly opposite to 

 the general principles of Cope's assumptions, in that slow meta- 

 morphosis at low temperatures reduces the larval foot and pro- 

 duces' ver}-- slender round toes, and this is especially accentuated 

 if the larva was already a slender-toed example grown at the low 



240 



