730 PROCEEDINGS OF THE AMERICAN ACADEMY. 



The glass tubing varies from 5 or G mm., inside diameter, to 7 or 

 8 mm. outside. The inside of the tube should be thoroughly cleaned 

 before use. For this purpose a bundle of tubes is entirely immersed 

 for some time in concentrated sulphuric acid to which sodic chromate 

 has been added. The tubes are then washed, dried, and stored away 

 from dust. In drawing out the capillary, care should be taken to draw 

 to the same outside diameter and as nearly as possible to the same 

 length. If one starts with tubes of as nearly equal size as possible, 

 and observes these conditions, a capillary of nearly constant bore is 

 obtained. It is on this uniformity of bore that the gradation of the 

 standards depends. One should therefore at least draw the tubes to 

 the same diameter, since this condition is easily governed and is the 

 chief factor in determining the bore. We have used for this purpose a 

 Brown and Sharpe wire gauge, and have drawn the tubing to gauge 

 No. 13, which corresponds to 1.8 mm., or 0.072 inch. 



In beginning the run, three to five grams of zinc are placed in the 

 reduction flask, which is then attached at one end to the stopcock of 

 the constant generator, at the other to the drying tube. To the latter 

 is then attached the ignition tube, with the collar over the capillary. 

 20 c.c. dilute hydrochloric acid are added through the funnel tube, 

 and the apparatus may be tested for tightness by closing the end of 

 the capillary with the finger while adding the acid. The hydrogen 

 is then turned on from the generator, and, when the air is expelled, is 

 lighted at the end of the capillary and turned down to a height of 

 about one millimeter. It is important that this height be maintained 

 as nearly as possible throughout the run and that the flame should 

 burn steadily, since an irregular flow of hydrogen results in an uneven 

 deposit of antimony. After the apparatus has been in action for five 

 or ten minutes, the measured amount of antimony is added to the 

 reduction flask through the funnel, which is then rinsed into the, flask 

 with a little water. No air is introduced into the apparatus if the 

 funnel is sufficiently small compared to the tube in which it is set, and 

 if the lower end of the tube is constricted. 



The deposit of antimony makes its appearance in five to ten min- 

 utes, and is completely deposited in thirty, though we have usually 

 waited forty to fifty minutes, to avoid the possibility of loss. We have 

 satisfied ourselves that all the antimony is deposited from the hydride 

 at this point, since a further heating along the tube gives no mirror. 



The tubes containing the mirrors are sealed at each end and 

 mounted in a frame, as shown in the plate. This firame is 185 mm. 

 by 70 mm. outside, 135 by 35 mm. inside, and made of blackened 

 wood G mm. in thickness. The tubes are fastened in holes passing 



