8o 



OPTICAL PROJECTION 



of brass pipe, is also brazed. Over the open bottom end i<? 

 stretched half of a piece of stout vulcanised rubber tube, (DD), 

 about two inches long. The other end of this rubber stretches 

 tightly over the ordinary size of any narrow-necked bottle, 

 using when at home Winchester quart size. When from 

 home, however, ordinary wine or ale bottles can be used, and 

 thus only the small fittings need be carried about. 



When only two bottles are used, the second should only 

 be half full, in order that as little spray as possible may pass 

 away with the gas. Still the latter 

 will be damp, and should be further 

 dried ; and for years I have used, after 

 the second purifier, a piece of large 

 glass tube, corked at the bottom, and 

 over the top of which stretches another 

 of the fittings shown in fig. 49. The 

 entry pipe goes nearly to the cork, and 

 most of the aqueous vapour is there 

 finally deposited. 



This is what is meant by properly 

 D purified and dried oxygen gas. If the 

 two bottles and drying-tube are ar- 

 ranged on a little shelf, or in a box, 

 with their connections ready made, 

 such a really efficient apparatus gives 

 no trouble whatever, beyond renewing 

 the alkaline solution now and then. 

 Should the third or drying chamber 

 not be employed, the gas can be 

 tolerably dried by placing the gas-bag 



which is being filled upon a table, while the rest of the 

 apparatus is below on the floor ; a great deal of the moisture 

 will then be deposited in the tubing connecting the bag. 



Such being the detail of the process, a few words will be 

 useful as to its general conduct. It is convenient to make a 



\ 



FIG. 49 



