( 454 ) 



in this case .4 lias to be jmt in the same relation to B iu which that respective 

 representative stantls to B, i.e. it must be considered cither as specifically different 

 or as specifically identical with B, according to the specific distinctness or uou- 

 distinctness of that representative. 



The same kind of evidence we may employ when we have to come to a decision 

 as to the specific distinctness of geogra])hically se])arated forms wliich are not 

 connected by intergradations. But when that endence is not conclusive enough, we 

 may have recourse to the evidence furnished by the variation of the forms. We 

 must accept as a general law that forms which are connected by all intergradations, 

 or forms which overlap in characters, are specifically identical ; geogra])hical form, 

 agreeing with this law are, therefore, to be accei)tcd as specifically non-distinct. If 

 we now compare the various organs of the species in respect to the effect which the 

 causes of variation have upon them, we shall find that a number of characters are 

 easily affected and show a variation between wide limits, while other characters 

 remain comparatively constant. Organ a varies, for example, in sjiecies A, as far as 

 we know at the time, from ten to a hundred, while organ b varies only from thirty to 

 thirty-five. Now if it is proved that in a nimiber of allied species a similar difference in 

 respect to the variability of the organs a and h takes ])lace, we can with great proba- 

 bility of correctness conclude that a form B similar to A is specifically distinct from 

 .1 if the character of the organ h is far outside the range of variation observed in A, 

 and, on the other hand, that jB is a form of the species A if the character of the 

 organ b comes within the limits of variation observed in -4, no matter whether A 

 and B are very similar or very dissimilar in respect to the variable character of the 

 organ a. 



As long as the special evidence does not force us to conclude otherwise, the 

 diagnosticist has to go by the following two general rules : — 



1. If is found that A and B stand in a certain relation to one another (sexes, 

 aberrations, seasonal forms, subspecies, sjiecies), and that the allied forms (' and D 

 differ from one another in a similar way as A does from B, C and D have to be 

 put into the same relation to each other in which A stands to B. 



Illustration. — We know by breeding ex])eriments that in Japan the spring brood 

 of Papilio sarpedon is smaller and has a wider band than the summer broods : in 

 North India we find an insect very similar to the Japanese one, and observe that in 

 April and May a form flies which is small and has a wide band, and that later in the 

 year all the specimens belong to another form which is somewhat larger and has a 

 narrower band : as in Japan the smaller and the larger forms are proved to be 

 seasonal forms of one species, we are logically bound to regard also the smaller and 

 the larger forms in India as belonging to the spring and summer broods respectively 

 of one species. A good number of allied Papilio.i show in India the same 

 phenomenon ; specimens collected during the first lialf of the year are smaller and 

 have wider bands than the specimens collected later on ; thongli it has not been 

 proved by rearing that we have here actually to do with spring and summer forms, 

 it would be illogical to regard the spring and summer specimens as specifically 

 distinct. 



2. If it is found that A and B stand in a certain relation to one another, which 

 relation is either proved by experiment or arrived at by general reasoning, a 

 specimen or specimens differing from A and j5 in a similar way as A does from B 

 have to be considered as a third form C standing in the same relation to A and B 

 as A does to B. 



