556 



THE AMERICAX XATUEALIST [Vol. LIU 



and this, no doubt, Duerden is already doing. If the 42-plumed 

 cock has descendants with a higher plume number than 42, the 

 theory- will have been disproved, which would undoubtly be 

 highl}' pleasing to Duerden because it would give him a more 

 hopeful basis for economic work. Now my own experimental 

 work with loss-variations leads me strongly to hold the more 

 hopeful view, that genetic changes are plus as well as minus, even 

 in the case of structures which are in course of phylogenetic de- 

 generation. 



The degenerating lateral digits of the guinea-pig's foot' pre- 

 sent a case parallel with those of the degenerating wing and the 

 degenerating fourth toe of the ostrich. The gumea-pig, like all 

 wild species of the genus, Cavm, has lost altogether the first of the 

 five typical digits, and has lost the fifth digit from its hind foot, 

 but not from the front foot. Some years ago I discovered a 

 gumea-pig which had an imperfectly developed fifth digit on one 

 hind foot. Neither of its parents had a fifth digit on either hind 

 foot. This fact alone shows the possibility of plus fluctuation in 

 a degenerate organ. The polydactylous individual, a male, was 

 mated both with related and with unrelated females. By the 

 former, he had 18 polydactylous and :]2 normal individuals; by 

 the latter he had 2 ])olydactyh)us and -U) normal individuals. 

 ThiM.^ult >ho\\ed that noinial f<in<ih- nht.d , tti, pol\dactyl 

 male, even tiiougli tinnnselves nornnil. i I'aiisiiiil ml a factor or 



females, when hutli softs wciv maied -to the same polydactyl male. 

 Breeding tho poUdactx! ot1>[)ring together and continuing the 

 race by selecting llu;se indivitluals which had the best developed 

 toes (purely somatic selection), a race was secured within four 

 generations which produced regularly 90 to 100 per cent, of 



