64 



ANNUAL REPORT 1918 AND 1919 



be said that a preliminary study of the Hterature seems to indicate that we 

 are deahng with the Persea schiedeana of Nees, reduced by Meissner in 

 1864 to the rank of a botanical variety of P. americana. It seems that 

 very few botanists have collected this species, and that it is imperfectly 

 known botanically, the fruit having been seen by none of the botanists who 

 have attempted to classify the plant. Since ample botanical material is 

 now available its correct classification should be a simple matter, and I am 

 convinced from my observations of the tree and its habits that it is a 

 distinct species, not to be included as a variety of P. americana. 



The classification of the Fuerte and Puebla varieties is still in doubt, 

 as I have not yet been able to visit Atlixco to see the parent trees and 

 determine whether they represent a distinct race common in that region, 

 or whether they are aberrant forms probably due to crossing. It is to be 

 hoped that this matter can be settled before the termination of the investiga- 

 tions in Mexico. In the meantime, I do not believe these varieties should 

 be included in the Guatemalan race, as they evidently are not true Guate- 

 malans. This is indicated by the presence of the aromatic odor in the 

 leaves, a character which the Guatemalan race does not possess. 



In the past, numerous efforts looking toward a classification of culti- 

 vated avocados have proved unfruitful because the question was not at- 

 tacked from a sufficiently broad standpoint. The result has been a classi- 

 ficatory system which included nothing more than a limited number of 

 varieties belonging to one or more races. In glancing over one or two of 

 these attempts, one comes upon such terms as "round pagua," "large 

 pagua," "San Angel black," "Chalco green," and so on; these cannot be 

 considered natural classes or groups in any sense of the word. The terms 

 "round pagua" and "large pagua" would include in the one case all round 

 fruits of the West Indian and Guatemalan races, in the other all large 

 fruits of both these races, whether round or not. "San Angel black" and 

 "Chalco green" would include all black fruits from San Angel and all 

 green ones from Chalco. It happens that all of these belong to the Mexi- 

 can race, but under these two groups we would get a motley collection of 

 forms and sizes. The defect of the system is that it has no regard for 

 the race to which the fruits belong, but classifies them upon the basis of 

 form or color, a purely artificial arrangement and one that is only satis- 

 factory where a natural classification, — one based upon relationship and not 

 upon arbitrarily chosen characteristics, — is utterly impossible. 



If a classification upon an arbitrary basis is undertaken, using form, 

 color, and size of the fruit as classificatory characters, it must not be made 

 to extend beyond the limits of a race. Some such classification may be 

 desirable, later on, to bring together all varieties of similar characteristics 

 within each race. Thus all varieties of the Mexican race can be classified 

 upon the basis of form, size, season of ripening, or some other importzuit 

 characteristic. 



Distribution of the Races 



This subject can only be touched upon at the present time, since many 

 sections of Mexico have not yet been visited. It is possible, however, to 

 give an idea of the distribution of the various races and to mention the 

 regions which have the reputation of producing the finest fruits of each. 



