316 The Unity of the Organism 



vidual salpa, for example, ever produces its like. In some 

 species parent and offspring are very unlike, so much so 

 that were they not actually observed to be parent and off- 

 spring, they would be regarded as unrelated and belonging 

 to different genera. But instead of being wholly unique, 

 though so unlike its parent, the young returns to its grand- 

 parent for a pattern; so if we jump a generation the rule 

 holds after all. This scheme of reproduction, technically 

 known as alteniation of generations, occurs in a considerable 

 number of groups of both plants and animals. Another 

 exception is presented by the jNIendelian mode of heredity. 

 One of the most characteristic tilings about this kind of in- 

 heritance is the skipping of generations as regards heritable 

 attributes. When gray and white mice are mated the issue 

 are all like the gray parent ; but some of the grandchildren, 

 if inbreeding be followed, are like their white grandparent. 

 This departure from the rule of like is so important and 

 peculiar that some biologists have felt it necessary to frame 

 the definition of heredity so as to make it cover the appear- 

 ance in offspring of difference as well as of resemblance. In 

 truth though, if the term descent be understood to pass over 

 one or more generations, as in tlie case of Salpa, tlie rule 

 holds. Indeed, Mendelian inlieritance in hybrid races miglit 

 be described as a sort of alternation of generations. 



Another aspect of the law of like must be noticed here. 

 Not only do organisms come from ancestors but we have 

 not a scrap of trustworthy evidence that they are or ever 

 have been produced by any other means. In other words, 

 the law of biogenesis, the law, that is, tliat negatives the 

 theory of spontaneous generation, is the same in large part 

 as the law that like produces like. So starting from any 

 given individual, problems of genesis and resemblance look 

 in two opposite directions, backward into the past, and for- 

 ward into the future. Viewing heredity from this standpoint 

 compels us to consider closely the degree of resemblance 



