545 
to face with the old dilemma. Do these circumstances 
produce a change in the organism by their direct trans 
forming influence, or are they the results of Natural 
Selection ? The case of the proteus just quoted goes to 
show that there is a direct response. The organ which is 
being used for a time is enlarged, while the organ which is 
being unused grows smaller through inaction. 
But if circumstance and outward condition act in this 
way in the case of animals who are provided with both 
lungs and gills, why should they not exercise a similar 
direct influence over the merest rudiment of a lung and 
the fully-developed gill ? Assuming the necessity for 
breathing the air, why should not the rudimental lung 
increase in proportion as the fully-developed gill decreases ? 
But it does not seem to me to be equally easy to picture 
this transmutation taking place by means of Natural 
Selection. So long as there is an ample supply of air in 
the water which the fish inhabits, there is no reason to 
doubt that he would remain as he was. This hypothesis 
is confirmed by the experiment made on larval amphibians 
whose development was retarded by the oxygenation of 
the water in which they lived. 
Whatever the cause of the want of supply of air in the 
water, there is only one other possible way of supplying 
the demand, and that is to seek to secure a supply from 
the air of the atmosphere. Now, if the same necessity 
affected a group of a given species, and they all sought, at 
the same time, the same solution of the difficulty, they 
would all be subject to the transforming influence which is 
so powerful in the case of the proteus. 
In order that the evolution of the lungs should be 
brought about by Natural Selection, three things are 
necessary : the elementary organ must vary in each indi 
vidual by reason of the variations inevitably associated 
II