48 RUBBER-CONTENT OF NORTH AMERICAN PLANTS. 



4.4 per cent from the leaves and 2.3 per cent from the stems. This 

 plant was growing in alkaline soil, which fact may have influenced 

 the formation of rubber in its stems or perhaps have caused its trans- 

 location from the leaves. It is to be noted that in A. mexicana the 

 stems are green and doubtless share with the foliage in the photo- 

 synthetic activities of the plant. It is probable that a species of this 

 sort would respond more readily to experiments designed to increase 

 the rubber-content of the stalks than would one in which the stems 

 played little or no part in photosynthesis. 



The exceptional species referred to above, in which the stems carry 

 more than a negligible percentage of rubber, is Asdepias subulata, a 

 milkweed of the southwestern deserts and Lower California. In this 

 plant the leaves are so few and small that it has not been possible thus 

 far to assemble a sample for analysis, but the stems themselves take 

 over the r61e of leaves to a considerable extent and in consequence 

 are found to carry considerable amounts of rubber, in some specimens 

 as high as 6 per cent. It is thus again suggested that the formation 

 of rubber is closely connected with the photosynthetic activity of the 

 plant. This in turn lends support to the view that the formation of 

 caoutchouc is closely associated with that of carbohydrates. In this 

 connection it should be noted that the bark of the stem in A. subulata 

 carries much more rubber than the stele itself (No. 957 of table 1). 

 This suggests that there is a prompt transformation of the carbo- 

 hydrates if the caoutchouc is a conversion-product of the sugars, as 

 investigators of the chemistry of rubber formation are now inclined 

 to believe (Harries, 1905, 1919). 



The almost total absence of rubber in the old woody parts, and 

 especially in the roots, is expected in latex plants. That this holds for 

 Asdepias is shown by a number of analyses of these parts (for example, 

 No. 1015 of table 1, and No. 304 of table 9). 



Variation due to heredity and to environment. During the three 

 years' time devoted to the studies here described, much attention has 

 been given to the extent and cause of the variation in rubber-content 

 of the species of Asdepias. With all of the data now assembled, the 

 only positive statement that can be made is that there is a wide fluc- 

 tuation both between the 16 species examined and within each of the 

 several species studied in detail. In Asdepias sullivanti, which has 

 received more attention than the others, the content was found to 

 range from 1.2 to 8.1 per cent in the leaves and 0.2 to 0.9 (in one anom- 

 alous case 8.2) per cent in the stems. In case an attempt is made 

 to improve the plants for commercial purposes, it will be important to 

 determine whether such variations are due to differences in the herit- 

 able properties or to environmental causes. A start has been made 

 on this problem, but it is so baffling that no positive conclusion has 

 been reached. Some of the results will be indicated a little farther 



