THE RESERVE MATERIALS OF PLANTS. 75 



nasturtium Tropceolum ma/us, and took the form of an 

 investigation of the nature of their carbohydrate con- 

 stituents under various conditions of illumination or in- 

 solation. A number of leaves detached from the plants in 

 early morning were divided into two portions; one set were 

 killed at once by heating in a steam chamber ; the others 

 were exposed with their petioles in water to sunshine for 

 twelve hours, being spread upon wooden trays covered 

 with wire gauze. They were then killed as were the former 

 set. The contents of the cells showed great difference in 

 their carbohydrates, the starch increasing from 1*23 to 3*91 

 per cent, of the dry weight, and cane-sugar going up from 

 4*65 to 8*85 per cent. In another experiment the increase 

 was for starch from 3^24 to 4*22 per cent., and for cane- 

 sugar 4*94 to 8'02 per cent. At the same time the total 

 sugars of the leaf, including laevulose, dextrose and maltose, 

 nearly doubled in amount. The chief feature which thus 

 accompanies assimilation of carbonic dioxide in these leaves 

 is the formation of a great amount of cane-sugar, the 

 absolute quantity being always much greater than that of 

 the starch. The authors conclude that when assimilation 

 is vigorous this body is made to accumulate much faster 

 than it is used, and the leaves are consequently charged 

 with it, possessing therein a temporary reserve material of 

 carbohydrate. As soon as the concentration of this sugar 

 in the cell-sap and protoplasm exceeds a certain strength, 

 starch commences to be elaborated by the chloroplasts, 

 forming a more stable and permanent reserve material than 

 the cane-sugar. The smaller increase of the starch as 

 compared with that of the cane-sugar is intelligible on the 

 hypothesis that the latter, and not the former, is the first 

 constructed carbohydrate, and that the starch is formed as 

 a secondary product from it, as in the case of underground 

 deposits it is produced by the leucoplasts from another 

 sugar transported to them. 



On reviewing these experiments it at first seems strange 

 that cane-sugar should be the immediate antecedent of 

 starch, the latter being so much nearer maltose in its con- 

 stitution. The conclusions are strengthened, however, by 



