274 Rapid Clinical Test for Hyperglycemia [Jan. 



in the proportions employed because i c.c. was not sufficient to give 

 a clear Solution. The addition of the extra c.c. of sulfuric acid sol. 

 removed this difficulty. 



Equally satisfactory results were gotten with diluted normal 

 human blood, artificial hyperglycemic human blood and diabetic 

 blood. The blood was made artificially hyperglycemic by the ad- 

 dition of sufficient glucose to make the proportion 0.2 per cent. 

 The difference between the normal and the two hyperglycemic 

 bloods, in their response to the test, was very marked with the use 

 of I drop of the diluted blood, more marked with 2 drops. 



An attempt was made to compare the responses to the test by 

 blood with those by sugar. This was found to be impossible be- 

 cause the color produced by blood is different from that caused by a 

 sugar Solution. By substituting serum for blood, the colored Solu- 

 tions obtained with the Molisch reagents approximated far more 

 nearly to that obtained with a pure sugar Solution. In this way, by 

 comparison with Standard sugar Solutions, quantitative measure- 

 ments were undertaken. However, the results were invariably high 

 (0.1-0.2 per cent for normal, and from 0.3 per cent up for hyper- 

 glycemic samples). This may be attributed, in part at least, to the 

 serum proteins, for, as is well known, proteins give a positive 

 Molisch test. Quantitative estimations, after removal of the pro- 

 teins, are now being attempted. 



Conclusion. The Molisch test is available for the detection of 

 sugar in blood. The difference in color between tests performed 

 on normal and hyperglycemic bloods is very marked and very readily 

 detected. 



