1615 
of the parents. As in viviparous Amphibians the jelly membrane 
is preserved, there is no question of direct gas-interchange between 
parental and foetal blood and respiration must take place between 
the air in the parental breeding cavity (or the parental blood) and 
the foetal blood passing through the jelly membranes and the liquid 
contents of the eggchamber, not exactly a short way ! 
If development takes place out of the water, but not within 
parental breeding cavities, precautions have to be taken to reduce 
evaporation, which necessarily will limit the otber interchanging 
processes. In all cases respiration is seriously impeded and in con- 
sequence of this circumstance respiratory organs increase enormously 
in size. External gills may act as such and acquire a much larger 
extent than in aquatic larvae (Gymnophiones, Salamandra atra 
Laur ete), but the tail also may grow in length and show a very 
abundant vascularisation, in order to facilitate gas-interchanging 
processes (Hylodes, Phrynivalis, Mantophryne ete), while in rare 
cases (Rana opisthodon Boul.) the whole skin and especially 
cutaneous folds on the side of the body supersede the other respi- 
ratory organs. We may conclude from these examples that widely 
different parts of the body may assume respiratory functions. Usually 
these organs possess at the same time a nutritive function the 
liquefying and absorption of the albuminous jelly membranes (hence 
the name allantoidean gills). 
Now in my opinion anamniote ancestors of Amniota may have 
been viviparous in a manner such as I have sketched above, but in 
this case the yolksack may have assumed a respiratory function. 
The fact that the yolksack is always abundantly vascularized in yolk- 
laden eggs, since haemopoiesis chiefly takes place on the yolksurface 
and bloodvessels play an important part in absorbing and carrying 
off the reserve-material to the embryo proper, favours this proposition. 
The yolksack will show the same tendency to increase its surface 
as the above mentioned respiratory organs, but the less plastic yolk- 
cells will oppose. The consequence will be that somatopleura and 
splanchnopleura of the yolksack separate, the first carrying off part 
of the bloodvessels (allantoidean vessels) and exclusively taking care 
of respiration, while the splanchnic sheet will retain the umbilical 
bloodvessels, bloodformation and yolkabsorption being restricted to it. 
The somatic sheet which may be called pneumatoblast, once free 
from the impeding yolkmass, will rapidly enlarge its size and try 
to increase its surface by folding-up and by proliferating. Between 
somato- and splanchnopleura an extensive cavity: the exocoeloma 
will arise. The pneumatoblast also will assume a nutritrive function : 
