PROCESS OF REGENERATION 113 



This is the typical annelid gradient, — i.e., one in which 

 the rate of metabolism decreases from the head backwards and 

 rises again at the posterior end.^ This gradient, as already 

 stated, is due to the characteristic method of growth of annelids 

 by formation of new segments posteriorly. If the posterior end 

 has been growing very rapidly, as it does when there is abundant 

 food supply, its rate of metabolism may be higher than that of the 

 head, and it may, therefore, disintegrate first, but the disinte- 

 gration of the head always follows shortly after. 



Three stages in the disintegration of Dero limosa are illustrated 

 in figure 8, and a graph of the same in figure 9. In this graph, 

 and in the succeeding ones, I have attempted to compensate for 

 the continual decrease in size of the posterior segments by gradu- 

 ally increasing the number of segments per unit of the cross- 

 section paper. In this way, a truer picture of the gradient is 

 obtained. 



The posterior disintegration may begin with the anal segment, 

 or in the region just anterior to this where the youngest segments 

 are forming. In the latter case, the anal segment, which in the 



' As a matter of fact, Morgan ('04), if he had only known it, long ago demon- 

 strated the typical annelid gradient in the earthworm by means of a galvanometer. 

 He found that in the earthworm the anterior and posterior ends are electronega- 

 tive to the middle part. It is a familiar physiological fact that stimulated regions 

 (i.e., regions of increased metabolism) are electronegative to non-stimulated ones 

 (current of injury, negative variation, electrical variation during the heart 

 beat, etc). The fact that cutting is a stimulation also accounts for the general 

 result of Morgan that cut surfaces of the earth worm are electronegative to 

 intact ones. The results of Morgan also indicate that the clitellum and the 

 fifteenth segment are local regions of high metabolic activity, probably owing 

 to their secretory nature. The data which seemed inexplicable to Morgan there- 

 fore find easy explanation when the nature of the annelid gradient and the stimu- 

 lating effect of cutting are known. There can be no possible doubt that the 

 difference of potential between the intact and the cut surfaces has a different 

 value when the cut surface is to form a head than when it is to form a tail, but 

 there is no reason whatever for assuming that a reversal of potential should occur 

 in the tail forming region of the earthworm. In experiments of this kind there 

 are many important factors to be considered, such as the length of time after 

 cutting, length of the piece, part of the animal from which the pieces are cut, part 

 of the intact surface to which the other electrode is applied, etc. The obser- 

 vations of Czwiklitzer (Arch. f. Entw'mech, Bd. 19) on the disintegration and 

 death of the polychaete Ophyotrocha also constitute a demonstration of the 

 annelid gradient. 



