1909.] OF THE TREE-FROG PHYLLOMEDUSA SAUVAGII. 895 



distributed among the egg-mass, at the time of hatching the 

 relatively enormously distended vitelline membranes fill a far 

 greater bulk of the nest, and the jelly capsules between them are 

 reduced to an insignificant remnant. 



When about to hatch the tensely filled membranes burst at 

 the slightest touch, liberating both embryo and fluid. If a nest 

 is opened soon after the eggs have hatched, it presents a seething 

 mass of tadpoles wriggling about in a thick mucilaginous fluid, 

 formed by the clear liquid from the burst vitelline membranes 

 and the now dissolved remains of the jelly, in the interior of a 

 chamber the sides of which are formed hy leaves, and the floor and 

 roof by the plugs of empty capsules. 



In order that the larva? should reach the water beneath them 

 it is necessary that the wall of this chamV)er should give way 

 somewhere. The fluid above has a softenins' effect on the g-ela- 

 tinous floor of the nest, and this gradually softens. At a period 

 of about 12-24 hours after the bulk of the larvpe are hatched 

 (there seems to be about a day's interval between the hatching of 

 the first and last larva) a thick mucilaginous di'op may be seen 

 to form at the bottom of the nest, and presently there is a steady 

 drip of the deliquesced jelly plug into the water below. After 

 a few minutes a larva slips through and falls into the pond beneath. 

 A few seconds later two or three more come through in the same 

 way and then they come faster and faster as the whole semi-fluid 

 contents of the nest continues falling drop by drop into the water, 

 taking the larva? with it. One nest, in which I watched the 

 whole process, took five minutes to empty itself, in which time 

 over 300 tadpoles fell from it into the water. 



It sometimes happens that a nest is hung a few inches from the 

 edge of the water. In this case the tadpoles suffer no inconveni- 

 ence from falling on the dry earth, but being extremely agile 

 quickly fiick themselves into the water. Budgett mentions this 

 happening in P. h^pochondrialis also. 



The larvfe, like the aquatic young of so many other vertebrates, 

 exhibit a retraction of their chromatophores at night and an 

 expansion in the daytime. 



The most interesting feature of this process is the part played 

 by the empty egg -capsules, which may be said to be three-fold. 



First, — The plugs at the top a,nd bottom of the nest provide 

 shields from the sun and air for the eggs, where the leaves do not 

 protect them. The eggs are quite unpigmented, and any that are 

 exposed to the surface, as happens often in less perfectly formed 

 nests, turn yellow and die. 



Secondly,— The empty capsules mixed with the full ones in the 

 body of the nest provide an extra source of fluid for the developing 

 embryo, and for the newly hatched larva, as already described. 



Thirdly, — The plug at the bottom sei'ves to keep the whole nest 

 intact, until the rather diffuse process of hatching is completed, 

 and all the larvae are ready to fall into the water. 



The large number of empty capsules mixed with the full ones 



