loo SECTIONAL ADDRESSES 



investigation by the systematist, as the result may be quite interesting 

 and perhaps important for our understanding of locaUsed differences 

 encountered among other animals. 



Systematics, however, are not concerned with the study of species and 

 their variations only. The species have to be grouped into genera and 

 then into higher categories, all according to relationship — i.e. according to 

 descent. As in the study of subspecies the systematist must enter upon 

 geography, so in the search for the past connections between genera and 

 families his research becomes linked with the past history of the Earth 

 and sometimes throws light on this history. If he can prove that two 

 genera now widely separated geographically are really of common stock, 

 then there must have been a means of communication in former times 

 which is now absent. If I may draw again on my brother's studies for 

 an illustration, we will take the distribution of the queerest-looking fleas 

 as yet discovered, the Australian Stephanocircus and the American 

 Craneopsylla, in which the anterior portion of the head is divided off as 

 a laterally compressed helmet. They are closely related, and the group 

 originated in South America, where occur several allied genera and a 

 genus connecting the group with more normally built fleas. They are 

 only found in the Andesian countries from Patagonia to Ecuador (possibly 

 occurring farther north), and in a modified form as Stephanocircus in 

 Australia, nowhere else. The assumption that there was at one time a 

 bridge between South America and Australia is the only explanation 

 at all satisfactory. This conclusion is supported by another genus (or 

 group of genera), Parapsyllus, which is plentifully represented by species 

 in the same Andesian countries (not in Eastern Brazil, the Amazons and 

 Guianas), and recurs in one species on the islands in the South Polar Sea 

 and in southern districts of Australia. The distribution of both genera 

 evidently took place from West to East. To this example, affording 

 positive evidence of a geographical bridge of some kind, may be given a 

 comparison which is a negative witness. One of the most remarkable 

 lacunae in the butterfly fauna of Africa south of the Sahara is the total 

 absence of swallowtails which feed as larvae on Aristolochia. The species 

 are numerous both in America, especially in the tropics, and in the 

 Oriental region, but not a single one has reached Tropical and South 

 Africa, though food plants occur, only one species of Ceylonese affinity 

 being found on Madagascar. In face of this evidence it is impossible to 

 believe that after the appearance on Earth of the butterflies there ever 

 existed a bridge between Africa and South America. 



Although the systematist is primarily concerned with the organisms 

 as produced by Nature, and not with the creative forces which have 

 evolved them, his researches extend to so many different species that he 

 is bound to collect evidence bearing on those forces and their working. 

 There are, in fact, certain questions which can only be answered with the 

 help of extensive systematic collections : Convergent development, for 

 instance, which looms rather large in discussions on natural selection, 

 particularly its frequency and its geographical occurrence. It is a fairly 

 common phenomenon, which however I shall mention only in passing, 

 for similarity in colour — such as shown in the mountains of New Guinea 



