REPRODUCTION. 933 



racteristics in castrated males, and of male characteristics in females with dis- 

 eased ovaries or after the end of the normal sexual life. 



Latency may be offered as the explanation of the numerous cases of 

 atavism, or reversion, by which is meant the appearance in an individual of 

 peculiarities that were formerly known only in the grandparents or more 

 remote ancestors, but not in the parents of the individual. This subject is one 

 of the most important in the whole field of heredity. Almost any character 

 may reappear even after many generations. In the human species stronger 

 likeness to grandparents than to parents is a frequent occurrence. The 

 majority of the frequent anomalies of the dissecting-room are regarded as 

 reversions toward the simian ancestors of the human race. The crossing of 

 two strains develops a strong tendency to reversion, and because of this the prin- 

 ciple of atavism must constantly be taken into account by breeders of animals 

 and growers of plants. As an example of reversion after crossing may be 

 mentioned the well-known one, studied by Darwin, of the frequent appear- 

 ance of marked stripes upon the legs of the mule, the mule being a hybrid 

 from the horse and the ass, both of which are comparatively unstriped but 

 are undoubtedly descended from a striped zebra-like ancestor. Here the 

 capacity of developing stripes is regarded as latent in both the horse and the 

 ass, but as made evident in the mule by the mysterious influence of crossing. 

 Darwin thinks likewise that the customary degraded state of half-castes may 

 be due to reversion to a primitive savage condition which, usually latent in 

 both civilized and savage races, is rendered manifest in the offspring that 

 results from the union of the two. Reversionary characters are often more 

 prominent during youth than during later life a fact that has been quoted in 

 favor of their explanation on the theory of latency. 



Regeneration. The facts of regeneration of lost parts must also be taken 

 into account in a theory of heredity. Such regeneration may be either physi- 

 ological or pathological. Physiological or normal regeneration has reference 

 to the reproduction of parts that takes place during the normal life of the 

 individual, such as the constant growth of the deeper layers of the epidermis 

 to replace the outer layers that are as constantly being shed. Pathological 

 regeneration refers to the replacement of parts lost by accident, and presents 

 the more interesting and striking examples. The power of pathological 

 regeneration in man and the higher mammals is limited. A denuded surface 

 may be re-covered with epithelium ; the central end of a cut nerve may grow 

 anew to its termination ; the parts of a broken bone may reunite ; muscle may 

 reappear ; connective-tissue, blood-corpuscles, and blood-vessels may develop 

 readily; and in the healing of every wound a regeneration of parts takes 

 place. But in descending the scale of animal life the regenerative power 

 becomes progressively stronger, and in many plants and low animals it is 

 marvellous. Thus, the newt may replace a lost leg, the crab a lost claw, the 

 snail an eyestalk and eye. If an earth-worm be cut in two, one half may 

 regenerate a new half, complete in all respects. A hydra may be chopped 

 into fragments and each fragment may re-grow into a complete hydra. From 



