Variation in the Food Reserves of Trees. 243 



tion occurring in spring. The roots of all nine trees studied in 

 the present work showed much starch in both phloem and xylem 

 all through the winter. 



6. The transformations of the carbohydrates are largely de- 

 pendent on the season, though the immediate conditions of tem- 

 perature have some effect. Thus, Russow and d'Arbaumont 

 found that several species kept in a warm glass-house over winter 

 lost their starch at the usual time; and in the work reported in 

 this paper, roots of trees exposed to the severity of winter by 

 removing their covering of earth did not appreciably reduce their 

 starch. On the other hand, it is known that a stem, without 

 starch in winter, will form starch in a few days after placing in 

 a warm temperature; and Russow reports several species of 

 trees that retained considerable starch in the stem through a mild 

 winter, but lost much or all of their starch in the next winter, 

 which was severe. 



7. Fabricius reports that the older stem of Picea excelsa does 

 not transform its starch to so great an extent as the younger 

 stems. Several of the trees examined in the present work have 

 shown the same thing; and hence it is quite likely that it is a 

 general phenomenon. 



8. Sablon has pointed out that the maximum for total carbo- 

 hydrate reserves for deciduous leaved trees is at the fall of the 

 leaf in autumn, whereas the maximum is at the opening of buds 

 in the spring for persistent leaved trees. 



9. Finally, the work of Sablon and Schellenberg indicates that 

 the principal carbohydrate reserve of trees in winter is cellulose. 

 One might wish that the methods employed by these two investi- 

 gators were a little more convincing. Sablon obtained favorable 

 results by chemical analyses, and both he and Schellenberg report 

 seeing the walls thinned down in spring ; in other cases the walls 

 showed a loss of refractive properties in cases in which reduction 

 in thickness could not be seen. There need be nothing suspicious 

 in the claim of a loss of substance where loss of thickness could 

 not be seen ; for enzyme action produces exactly this effect at the 

 first solution of walls in some seeds^. 



^Newcombe. Cellulose Enzymes. Ann. Bot. XIII, 1899, 49. 



16 



