198 ANNUAL OF SCIENTIFIC DISCOVERY. 



do not form a precipitate on boiling, except in the case of 

 glucose. 



Guided by these results, we are induced to recommend, for the 

 detection of glucose in urine, the following process, which cannot 

 give rise to any fallacy : 



Prepare the reagent by precipitating a solution of acid nitrate 

 of bismuth by a great excess of caustic potash, and pour a solu- 

 tion, drop by drop, into the moderately heated solution until 

 the precipitated hydrate of bismuth is completely redissolved. 



To recognize a diabetic urine, heat a portion with the above 

 solution. 



After a few minutes' ebullition, the urine becomes brown, and 

 metallic bismuth is then precipitated in the form of a black 

 powder of crystalline appearance, adherent to the glass, if glu- 

 cose is present. 



We have satisfied ourselves that the principles contained in nor- 

 mal urine, such as urea and uric acid, do not precipitate the 

 above reagent. Albumen only causes a brown color and a slight 

 turbidity, which we consider to be due to the formation of sulphide 

 of bismuth. 



Sulphuretted urines also give a black precipitate in a solution 

 of oxide of bismuth in potash and tartaric acid ; but this reaction 

 cannot be confounded with that caused by glucose. 



It is, besides, eas}^to recognize and (if desired) to separate the 

 albumen. Thus, on bringing to ebullition the urine of a person 

 suffering from Bright's disease, the liquid becomes turbid, opal- 

 escent, and deposits coagulated albumen. 



As to sulphides and sulphuretted hydrogen, these are easily 

 recognized by means of hydrate of lead, which these compounds 

 darken. London Chemical News, from Gazette Medicate. 



ACTION OF SUNLIGHT UPON GLASS. 



At a meeting of the Massachusetts Institute of Technology, 

 early in 18G7, Mr. Thomas Gaffield made a communication upon 

 the action of sunlight upon glass, giving the results of his experi- 

 ments, now extending over several years. He commenced his 

 ex-periments, then original as to their method and extent, more 

 than three years ago, and has prosecuted them to the present 

 time. Isolated cases of change of color in window-glass had 

 long been known, but they were attributed to some defect in the 

 manufacture ; his attention was directed to the subject by observ- 

 ing the change in glass of a very superior quality, lie began 

 by exposing several specimens of plate, crown, and sheet glass 

 during the month of July, and he found that a month's exposure 

 to the hot sun would change the best white French plate and white 

 sheet glass to a more or less yellow color; the dark-green shades 

 did not experience any change. Experiments for a longer period 

 confirmed these results, the hue becoming darker, according to 

 the time of exposure. That the color permeates the body of the 

 glass, and is not confined to the surface, or produced by reflection 

 therefrom, has been proved by grinding off about one-sixteenth 



