Mutations and Evolution , 
233 
limbs in diverticula of the branchial chambers, and such special 
adaptations as the ventral disc or sucker of certain mountain forms.” 
The existence of such adaptational characters is supported by the 
fact that certain species differ more widely in their tadpoles than in 
their adult stages. Thus (Boulenger and Annandale, 1918) the Indian 
species Rana tigrina and R . cancrivora are so similar as frogs that 
the latter was classed as a variety of the former. But Annandale 
(l.c.) has shown that the tadpoles differ markedly in buccal armature. 
Whether this is a case of convergence of the adult species, as 
Dr. Annandale thinks, or divergence of the tadpoles, as Dr. 
Boulenger believes, need not concern us here, though from analogy 
the latter interpretation appears more probable. 
To return to the recapitulatory characters, the transformation 
of the fish-like gill-arches of the tadpole into the aortic arches of 
higher vertebrates is too well known to require comment here, except 
to point out that it comes about through the gradual substitution 
of one series of blood vessels for another, the branches to the gills 
being gradually pinched off and the blood stream diverted to the 
more direct route to the lungs. The specious argument that the 
development of any recapitulatory character must go through such 
preliminary stages for purely structural developmental reasons is 
now seldom heard and can very well be consigned to oblivion. 
Developmental mechanics as well as comparative embryology tell 
strongly against it. 
The experiments of Gudernatsch (1914) in greatly retarding or 
hastening the time of metamorphosis by feeding tadpoles on thymus 
or thyroid respectively, showing that growth and differentiation are 
separate factors, do not affect our present interpretation. They 
merely indicate that the processes of development and metamorphosis 
are physiologically controlled by something in the body of the nature 
of hormones or enzymes secreted by certain tissues. A recent 
paper (Morse, 1918) has attempted a further analysis of the 
processes that lead to atrophy of the tail in metamorphosis. The 
writer concludes that autolysis is the primary physiological factor. 
The first step in atrophy, according to Barfurth, is the growth of 
the pygostyle which, by occlusion of the blood vessels in the tail, 
causes an accumulation of CO a and acids. This acidosis of the 
tissues induces autolysis. In this condition the phagocytes are 
chemotactically attracted to the atrophying organs, so that 
phagocytosis is a secondary, and not the primary factor as 
Metschnikoff supposed. Hormones or enzymes probably stimulate 
