2it0 The CarlKihiidvates of the MaiKjold Leaf 



sugars was much lower, viz. about 16 per cent. It is also striking that 

 the successive rises and falls in the proportion of total sugars at night 

 occur along practically straight lines (a phenomenon also observed in 

 Series II, although the curve of total sugars is not shown in Fig. 5). 



Fluctuation of the -^ niiio. 



In the final stage of growth in October, although the general type 



of the sugar curves is quite similar to that of Series II, the curve showing 



i.s. . 

 the ratio " is of an entirely different character. Whereas in Series I 

 c.s. -^ 



and II, during the greater part of the day, the curve showing the ratio 



of the two sugars runs very nearly parallel to the temperature curve, 



in the last stage of growth the curve shows a periodic increase and 



decrease of the ratio which the two sugars bear to one another. From 



9 a.m. to 11 a.m. saccharose increases faster than hexoses, from 11 a.m. 



to 1 p.m. the reverse is true; then comes a rapid increase of saccharose 



and a fall of hexoses until 3 p.m., which is reversed between 3 and 5 



p.m. At night there is a similar periodic fluctuation, and, exactly as 



in Series II, when the cane sugar increases at night it does so far more 



rapidly than the hexoses, cane sugar apparently being the primary 



product which is formed from the reserve substance, and not hexoses. 



On the other hand, exactly as in September, the saccharose falls ofl, from 



3 a.m. to sunrise, considerably more rapidly than the invert sugar, so 



I s 

 that the curve -^ rises steadily; as in Series II, the rise is almost along 

 c.s . 



a straight line. This is perhaps due to the inversion of the cane sugar 



being effected by invertase more rapidly than the invert sugar formed 



is consumed by the changes to which it is subjected. It is noteworthy 



that these changes, in this case, take place when the external temperature 



is below freezing-point; at 3 a.m. ice was forming on the leaves and at 



7 a.m. they were all frozen stiff. 



Pentosans, leaf-matter insoluble in Alcohol ami Pentoses. As in 



Series I and II, the curve of matter insoluble in alcohol is parallel 



to the pentosan curve. At this stage of growth, however, the variation 



during the day is much less marked than in September (Series II), the 



insoluble matter ranging from 4.5-48 ))er cent., the pentosan from 



(J-2 to (3-9 per cent. ; the greater part of the variation seems to be purely 



relative, that is the apparent fall of insoluble matter between 9 a.m. 



ami 1 I a.m. corresponds with the large increase of sugars (15-8 to 18-8 



