1612 
and attach them to leaves or stones, in this way preventing or 
reducing evaporation (Phyllomedusa, Hylodes martinicensis Tschudi, 
many species of the genus Rhacophorus, Rana opisthodon Boul. etc). 
In the third place the synoion may be deposited in damp, subter- 
ranean holes, the female curving around it for protection (many 
members of the family of Gymnophiones as: Siphonops, Hypogeophis 
Ichthyophis etc). 
In all these cases even in so-called viviparous forms the embryo 
remains surrounded by tertiary eggmembranes of colloid consistancy, 
the internal one liquefying in order to form an eggchamber, or by 
the fusing of many eggchambers a central cavity, filled up with an 
aqueous liquid. In this simple way Amphibians, which do not lay 
their eggs in the water, provide liquid surroundings for the developing 
embryo. It need not be said that these deviations from the normal 
way of development with a long larval period must be considered 
in most cases as the consequence of or as an adaptation to a non- 
aquatic life of the parents (usually to a burrowing life: Gymnophiones, 
and to an arboreal life, in the other examples) '). 
Wherever free larval life is supplanted by an embryonic period 
within tertiary jelly-membranes, we may remark two characteristic 
phenomena; the presence of a large amount of yolkmaterial within 
the egg and the enormous development of respiratory organs. Both 
phenomena may be considered as necessary adaptations to the changed 
conditions of life of the young Amphibians. 
If an aquatic larva capable of nourishing itself, is set free after a 
short embryonic period, the egg does not need to contain much 
reserve-material. If however development takes place wholly or for 
the greater part within tertiary eggmembranes, provision must be 
made for a more extensive victualling of the developing embryo. 
This explains the fact, so strange at first sight, that viviparous, ter- 
restrial forms as Salamandra atra Laur. (and to a certain degree 
also as S. maculosa Laur.) possess eggs much heavier laden with 
yolk than the closely allied aquatic Tritons, which deposit their 
eggs into the water. 
The considerable development of respiratory organs may be easily 
explained in the same way. In forms with a short embryonic period 
which is also passed, it is true, within tertiary eggmembranes, but 
where the greater part of the developing period is spent in the form 
of a free aquatic larva, the difficulties with which respiration meets, 
are far less serious than in forms where development takes place 
out of the water sometimes within folds, holes or internal cavities 
1) An exception must be made for Pipa, a thoroughly aquatic frog. 
