836 



THE AMERICAN NATURALIST 



[Vol. XL 



endemic or permanent, but occurs in epidemics at intervals of 

 from ten to twenty years. Its geographical distribution in South 

 Africa is very wide : in Natal, Zululand, the greater part of Rho- 

 desia, Bechuanaland, and Portuguese East Africa. Horses placed 

 in fly-proof shelters even in exceedingly unhealthy places in no 

 case incur the disease. The particular fly or insect carrier is still 

 unknown. As in several of the foregoing diseases the infective 

 power of the blood persists for years. 



Natural Origin of Immunity.— For the student of extinction 

 an important point to note, in connection with 'horse sickness,' 

 is that while artificial immunity is thus far undiscovered, degrees 

 of immunity and of natural immunity sometimes occur. Such 

 variations in respect to immunity would in a state of Nature lead 

 to the gradual selection of immune forms and the production of 



Summary as to Xafural Extinction bij Disease 



To summarize these remarkable conclusions which we owe to 

 the labors of Lewis, Koch, Theiler, Kilborne, Smith, Watkins, 

 Pitchford, and many others, we undoubtedly have here an agency 

 which must be seriously considered as an (H <"isi()nal if not a fre- 

 quent cause of extinction of (|nadru|)c<ls in the paM. It will be 

 noted (1) that in the case of the tsc-t^c fly .li.,.a^e the wild ru- 

 minants are the permanent though unharmed reservoirs of the 

 infective protozoan; (2) that in Texas fe\er or red-water fever 

 native immune Bovid* are the permanent carriers of the disease 

 organism; (3) that the 'rinderpest' ap})ears to he in an early stage 

 of its history as a disease in which neither .loniestieated nor wild 

 Bovidae have become naturally iininune and all the I5ovi(he act 



larly au-ain thai in • hor^e .iekn'es." ,,f South Africa the infeetive 

 ^ Thus in these .hsea^es we have all the conditions' favorable for 



