244 



Univrrslhj of Califoniid I'lililicatioiis in Botanji f^'"'- 



SuMMAitv or 'I'ahm: (i 



12 naiisoosus (typical) 



Ill making' use of tliis table it is bettor to take into account only 

 the first five varieties since the others are not represented by a suf- 

 ficiently lar<i:e number of samples to rtnider the rc^sults dependable. 

 ''Individual variation'' is an important factor and it is only after a 

 laro:e number of samples from widely separated localities have been 

 analyzed that one is justified in drawing conclusions. 



Taking tlicii only the first five varieties it would seem tliat tlu-y 

 would stand, as to rubber content, in the order of virkhdus, consimilis, 

 rjnaphalodfs, speciosus, (jravcolens. While tliis may be of significance, 

 it is also to be noted that the first two are inhabitants of alkaline flats, 

 the next two belong to non-alkaline slopes, and the last (as far as our 

 samples are concerned) to only moderately if at all alkaline soil. It 

 may be. tlierefore. tliat an apparent parallel between r\d)ber content 

 and botanical varieties is. in fact, due to environmental factors. 



While, as just shown, tlici'i' is no direct evidence that the botanical 

 varieties represent hereditary units, eaeli witli a dift'erent capacity for 

 iiil)l)rr pi'odnetion, tliere is nothing, on tlic other hand, to indicate 

 that eacli vari<'ty may not itself be composed of several or even 

 numerous biotypes. If this is the ease, Ihen tlie separation of these 

 tlirough selection might lead to the discovery of a superior strain. 

 The notable fluctuation in the rubber content of any one variety might 

 seem to indicate this possibility but the alternative as to the influence 

 of I'Mvii'diiincnt must also be kept in mind. 



