152 PROBLEMS OF RELATIVE GROWTH 



And in these the second and hinder horn is both smaller than 

 the first (and also less curved) . This implies that the growth- 

 gradient made visible in the form of the anterior horn is 

 continued across to the second horn-area, causing the growth- 

 intensity to diminish, and therefore resulting in a smaller horn ; 

 (and also that the shape of the gradient is not constant, but 

 flattens out, leading to less difference in growth-intensity 

 between the two ends of the horn-area, and consequently to 

 a decreased curvature of the second horn). These facts thus 

 lead to the same important conclusion as did the analysis of 

 the growth of the appendages of Eupagurus — namely, that 

 though intensive growth be restricted to specifically limited 

 areas (there the regions of the limb-buds, here the horn-areas), 

 yet the agency determining the growth-gradients, whatever 

 it may be, is organismal, and extends throughout the body. 

 It can only express itself where the potentialities for intensive 

 growth exist, but it is itself continuous (as in a rather different, 

 non-graded way, the hormones are distributed over the entire 

 system, but only exert effects where they meet with tissues 

 specifically adjusted to react to them). Thus we must assume 

 that even in the one-horned rhinoceroses, the growth-gradient 

 is continuous along the head, but can only reveal itself in 

 species where a second specific horn-area is present ; and 

 similarly, that in hermit-crabs the growth-gradient controlling 

 the relative growth of appendages is continuous, not merely 

 between the limb-producing areas of successive segments, but 

 even across large regions in which the capacity for limb- 

 production has been entirely lost, as in the anterior part of the 

 male abdomen. 



It is worth recalling that we already know of analogous 

 gradients, and to use a more general term under which gradients 

 can be included, fields, through the results of experiments on 

 regeneration and grafting. The mere fact that, normally, 

 precisely what is lost in amputation is restored in regeneration 

 points in this direction. The conclusion has been made more 

 probable by the proof given by Schotte, Guyenot and Weiss, 

 proof that in regeneration (of the Amphibian limb) the new 

 tissues are not, as was long held, proliferated from the old, but 

 are differentiated from a truly indifferent tissue, and will 

 differentiate normally even if the corresponding tissue has 

 been removed from the basal stump. And finally, it has been 

 clinched by the beautiful experiments of Milojevic, Weiss, 

 Locatelli, Guyenot and Schotte (references in Guyenot and 



