PROCESS OF REGENERATION 151 



ited heads; the higher the rate of /y the more will it inhibit the 

 anterior structure which arises, and if the rate of y is sufficiently 

 high, then region x will not be able to become independent at 

 all but will be dominated by y, and give rise to a subordinate part, 

 a tail. Under such conditions biaxial tails result. It is also 

 conceivable that x might begin head development but later be 

 dominated by y to such an extent that head formation can no 

 longer continue but tail formation sets in; this in my opinion is 

 the explanation of the cephaluran outgrowths. In the case of 

 multiple outgrowths, the outgrowth or outgrowths which have 

 the highest rate become heads, and dominate over the others 

 which then give rise to tails. In Planaria, when the region // 

 dominates the region x, the acephalic condition results; but this 

 is not the case in Lumbriculus, probably because the new tissue 

 begins to grow out before its fate has been decided, and it then 

 forms a head or a tail according as x or y dominate. 



It is not necessary that the region y should have a metabolic 

 rate which is actually higher than that of .r; in fact it probably 

 never has, for disintegration experiments show that x usualh' 

 disintegrates before y. But it is the ratio of the rate of x to 

 the rate of y which is important, and in all probability the rate 

 of x must exceed the rate of 1/ by a considerable amount before 

 normal head formation can occur. Re'atively slight alterations 



of the ratio . ^ — are sufhcient to affect the process of morpho- 



rate 01 y f i 



genesis. Thus is a set of pieces of the same size, and from the 



same level of the body, the value of y cannot be very different 



in the different individual pieces; the rate of x probably differs 



even less; yet from such a set of pieces all types of nterior 



structures are obtained. 



If this conception of the process of head formation is correct, 



an experimental control of morphogenesis should be possible. 



By depressing the rate of y as compared with x, one ought to be 



able to increase the percentage of normal heads; and, conversely, 



by increasing the rate of y as compared with x, a decrease in the 



percentage of normal heads should result. The first possibility 



is readily realized experimentally; the rate of y can be easily 



