BORIC ACID AND BORAX. 25 



be used in drying them, it was necessary to attach apparatus whereby 

 a current of air could be drawn through the can during the process of 

 desiccation. This was accomplished by soldering a tube to the side 

 of the can near the top and conducting the air, by means of a glass 

 tube shown in the section, to the bottom of th^ can. A second tube 

 was soldered to the opposite side of the can making connection with 

 the suction apparatus. The cans were of a magnitude to hold all the 

 dejecta in each case for a period of twenty-four hours. The weight 

 of the can being previously known, the weight of the moist dejecta 

 for the twenty-four hours was easily determined. 



For desiccating purposes the cans were arranged in a battery of six 

 for each series, though only five are shown in the drawing. In order 

 to collect any volatile nitrogenous or sulphur compounds, as well as to 

 determine the rate of passage of the air through the can, each one was 

 provided with washing bottles, as shown in the figure, through which 

 the emergent air bubbled on its course toward the vacuum pump. By 

 opening or closing the communication with the can the rate of passage 

 of the air was made uniform. Proper reagents were placed in the 

 washing bottles to hold back any of the volatile compounds above 

 mentioned. 



The vacuum was produced by a large aspirator connected with the 

 water service in the basement of the laboratory. As often twelve or 

 eighteen cans were operated at once it was necessary to have this jet of 

 rather large size. It was found that a jet of about one-fourth inch sup- 

 plied by a 2-inch pipe was entirely sufficient to produce a vacuum for 

 drying a series of three sets of cans as above arranged. The cans rested 

 in a copper tank filled with water to within about 2 inches of the top. 

 This water was maintained at a constant level by an automatic arrange- 

 ment of the usual description. The heat was applied by a series of 

 Bunsen lamps, so that all parts of the copper tank were evenly heated. 

 The temperature of desiccation was slightly below the boiling point of 

 water. At this temperature, and with the current of air regulated as 

 above described, the complete desiccation of the contents of the can 

 was accomplished in twenty-four hours. The loss of water having 

 been determined by reweighing the can, its dried contents were ground 

 and bottled for analysis. 



The construction of the can and desiccating apparatus is shown in 

 the accompanying figure (fi^. 1) in which a shows the cross section 

 of the can; , the water reservoir connected with the water supply, 

 arranged to fill automatically the baths used for drying the cans; <?, c, 

 lead pipes to supply air to the water reservoir for the purpose of per- 

 mitting the water to flow from it into the bath to maintain a constant 

 level; d, d, pipes to convey water from the reservoir to the bath. 



At first the ordinary tin paint cans were obtained from the manu- 

 facturers, and the side tubes soldered to them. It was found, however, 



