THE MOST IMPORTANT SPECIES. 19 



would expect from a casual examination of average stems and may 

 be accepted for purposes of general comparison. Only two samples 

 of the woody type have been analyzed. These carried less rubber than 

 the average for the stooling type, but the results might be quite 

 different if a large series could be examined. If the differences in 

 size and habit as above noted are purely ecological, it may be assumed 

 that only the stooling type would be obtained under cultivation, 

 regardless of the source of the seeds. If, however, these types repre- 

 sent hereditary races, then the selection of seed becomes a matter of 

 importance. In this latter case attempts should be made to hybridize 

 the forms in order to obtain a progeny combining the best characters 

 and with as high a rubber-content as possible. Judging from experi- 

 ence with other plant species, it is quite probable that the percentage 

 of rubber can be considerably increased through hybridization and 

 selection, and if each of the types already distinguished is itself com- 

 posed of minor strains, as is likely to be the case, these should first 

 be segregated in order to furnish material as a basis for the work. 

 An increase in the rubber-content is one of the requirements necessary 

 to bring such plants into the list of agricultural possibilities. 



The desert habitat of Asckpias subulata indicates that it may be 

 brought under cultivation more cheaply than other species. It cer- 

 tainly could be grown on land that is now 'considered worthless because 

 of the lack of water and which could therefore be had at a minimum 

 price. The low water requirement is evidenced by the growth made 

 under desert conditions. The normal rainfall at Sentinel, Arizona, 

 whence came many of the samples, is 4.2 inches; that at Yuma is 3.1 

 inches. The maximum temperature for 1919 at Sentinel was 120 F. 

 and the minimum 22 F. 



The better type of the desert milkweed grows in dry streamways 

 and in slight depressions on the desert mesas, indicating the desir- 

 ability of increasing the soil-moisture, either by slight irrigation or 

 through cultivation and other practices of dry farming. A system of 

 cross-furrowing or checking that would retard the run-off after rains, 

 together with surface tillage, would doubtless increase the tonnage. 

 Under such methods, and especially under irrigation, the plants would 

 probably develop normal foliage, but what effect this, or, for that 

 matter, the methods themselves, would have upon the total yield of 

 rubber can be determined only by experiment. 



It is highly desirable that cheap and effective methods of vegetative 

 reproduction be found for perennial plants to be grown for then* rubber. 

 Since Asckpias subulata grows from a taproot and is apparently lacking 

 in the horizontal, connecting root branches of most milkweeds, it 

 seems to propagate itself in nature only by seeds. However, some 

 recent and incomplete experiments made by Mr. H. R. Reed, while 

 connected with the Experimental Farm of the United States Depart- 

 ment of Agriculture at Bard, California, tend to show that vegetative 



