GLUCOSE. 153 



The sum of the variations in bath temperature was 5.57° and the 

 mean was 0.22°. The corresponding values for the parallel cane-sugar 

 series (II) were 4.35° and 0.21°, which shows that the success attained 

 in maintaining temperature was about the same in glucose Series I as 

 in cane-sugar Series II. 



The sum of the rotations of all the solutions used in glucose Series I 

 was 758.85°. The sum of all the losses was 8.60° or 1.13 per cent. The 

 dilution in the companion cane-sugar series was 2.86 per cent, or 2.53 

 times as large as in the case of glucose. 



The observed osmotic pressures have been corrected for all of the loss 

 in rotation, though, as explained in the preceding chapter, the dilution 

 which occurs when the cells are opened, if known, should be deducted. 

 But, since the total dilution was only 1.13 per cent, and since certainly 

 less than half of it occurred when the cells were opened, the results do 

 not greatly suffer by the inclusion of the latter. 



The striking features of Table 40 will be found in the last column, 

 in which are given the ratios of osmotic to the calculated gas pressures 

 of the solute. Considering the still undeveloped condition of the 

 method by which they were obtained, these ratios are remarkably 

 uniform throughout the whole series. The mean of all of them is 0.994, 

 and the greatest divergences from this mean are +0.013 and —0.013. 

 It will be recalled in this connection that, in the case of cane sugar, 

 the ratios of osmotic to gas pressure varied considerably from concen- 

 tration to concentration. The second noteworthy feature of these 

 ratios is that they approach unity — quite as closely probably as the 

 defects of the method at that time could be expected to permit. If 

 the approximate correctness of the pressures given in Table 40 is estab- 

 lished by later investigations, it will mean that, within the range of 

 temperatures 22° to 25°, the osmotic pressure of glucose solutions 

 obeys the laws both of Boyle and Gay-Lussac, since that is the only 

 interpretation of the unit ratios of osmotic to gas pressure. It is not 

 yet known whether this ratio will be confirmed for the temperatures 

 in question, since the work at 25° has not been repeated under condi- 

 tions insuring precision. It is already known, however, that at 30°, 

 40°, and 50° the ratio of osmotic to gas pressure is unity for solutions 

 of glucose. 



The molecular weight for glucose which is derived from the mean 

 ratio 0.994 — under the assumption that osmotic pressure obeys the 

 laws of Boyle and Gay-Lussac — is 179.82 instead of 178.74. 



In the case of cane sugar, Series I — without correction — gave a molec- 

 ular weight of 341.41 (0 = 16) instead of 342.22; while Series II— after 

 correction for the loss in rotation as inversion — gave a molecular weight 

 of 337.59 (H = l) instead of 339.60. The excellent molecular weight 

 which was legitimately derived from the results in glucose Series I was 

 partly responsible for the pertinacity with which, for a time, the mis- 



