THE LIFE CYCLE 55 



cases of regeneration, it is assumed that they contain either some 

 of the undifferentiated germ plasm or those elements of the germ 

 plasm which are necessary for the formation of the new part. Such 

 assumptions are not only vmsatisfactory because they cannot be 

 proved or disproved, but they are whoUy unnecessary. We have 

 seen that the organism can not only accumulate structural material 

 of various kinds, but under other conditions can remove to a 

 greater or less extent the material previously accumulated. Since 

 reduction occurs in organisms, we must at least admit the possi- 

 bility of dedifferentiation. Consideration of the data of observa- 

 tion and experiment is postponed to later chapters:' at present 

 only certain general features of the process need be considered. 



In the case of self -differentiation (see pp. 50, 51) the gradual 

 changes in the substratum may be reversed in direction under 

 altered conditions; the gel may again become a sol. But the 

 synthesis of new colloid molecules and the formation of new sol, 

 on the one hand, and the gradual breakdown and eHmination of 

 the old gel, on the other, is also possible. Apparently nuclear and 

 cell division are or may be factors in dedifferentiation. With the 

 occurrence of division the progressive changes in the cell, since the 

 preceding division, disappear more or less completely and the cell 

 returns to or approaches its original condition. An increase in 

 metaboHc rate is also apparently associated with division.^ If 

 the changes in one direction balance those in the other, cells which 

 divide may remain indefinitely embryonic, like the vegetative tissues 

 of plants and the growing regions of certain animals. But if the 

 nucleus or cell does not divide, or if division does not bring the 

 cell back to its original condition, then a progressive change must 

 occur in the cell or from one cell generation to another, and this 

 change appears sooner or later as differentiation and may go so 

 far that the cell finally becomes incapable of division. Where 

 differentiation has been a correlative process, isolation of a part 

 from the influence of the correlative factors which have determined 

 the course of its differentiation may result, if the part is capable of 

 reacting to the altered conditions, in metaboHc changes of such a 



' See particularly chap, v, and chap, x, pp. 245-47. 



' See chap, vi, pp. 141-42, and also Lyon, '02, '04; Spaulding, '04; Mathews, '06. 



