DIALTTIC SEPARATION OF GASES BY COLLOID SEPTA. 413 



silk, although not the most rapid in its dialytic action, was more convenient in use than 

 any other septum hitherto tried. 



The varnished silk stretched over a disk of porous earthenware (for support) closing 

 the mouth of the small glass bell-jar or osmometer, which has an opening of one 

 hundredth part of a square metre, gave 10 cub. centims. of dialyzed air in one hour, con- 

 taining 42*2 per cent, of oxygen ; barom. 767 millims., therm. 23°"5. For a square metre 

 of surface this is a passage of 2*77 cub. centims. of air per minute. 



A small bag, useful for experimental purposes, was constructed of a portion of the 

 same varnished silk, 053 metre in length by 0-27 metre in width, which had therefore 

 a surface of 0'143 square metre. The varnished side was turned inwards. Between the 

 folds of the silk was placed a double thickness of common felt carpet or a piece of 

 wadding, so as to occupy the interior of the bag. A glass quill tube also entered the 

 bag to the depth of a few inches, and projected as much outside, so as to admit of being 

 connected, by means of a sound adapter tube of French rubber, with a Sprengel pump, 

 as shown in figure 1 (page 408). The edges of the silk cloth were cemented round by 

 caoutchouc varnish, to a depth of 10 millims., so as to close the bag ; and care was taken 

 also to cement the glass tube well to the edges of the bag. When the silk bag is 

 exhausted of air, it remains nearly flat, and feels hard like a piece of cardboard. Such 

 an air-dialyzer is further improved by interposing a strong glass flask or bottle, of one 

 or two litres in capacity, between the bag and the pump, so that both are exhausted of 

 air at the same time. The flask must be strong enough to bear the full pressure of the 

 atmosphere without breaking. An auxiliary air-pump, to produce the first exhaustion, 

 cannot well be dispensed with where the space to be made vacuous is so considerable ; 

 the Sprengel tube is brought into action afterwards. The advantage gained by the 

 vacuous flask, and even by the thick wadding placed within the bag, is that they form 

 a magazine in which the dialyzed air can be allowed to accumulate for several hours or 

 a whole day, and from which the air may afterwards be drawn quickly by the Sprengel 

 tube for the purpose of experiment. A narrow glass receiver tube, which can be closed 

 by the thumb, may be used to take 6 or 6 cub. centims. for an observation on the 

 inflammation of a chip of wood in the highly oxygenated air. When the proportion of 

 oxygen is under 33 per cent, the wood is not rekindled; but in the ordinary action of 

 this dialyzer the oxygen is seldom found under 40 per cent. The best result is obtained 

 when the exhaustion is within half an inch of the barometric vacuum. When the 

 pressure was allowed to fall to one-half or one-third of an atmosphere, the proportion of 

 oxygen was lessened by 2 or 3 per cent. 



The action of heat and cold on the penetrability of rubber is considerable, as has 

 already been stated. Operating with the dialyzing-bag described, without any inter- 

 mediate flask, the volume of air collected in twenty minutes was 6-35 and 6-57 cub. 

 centims. in two consecutive experiments ; barom. 760 millims, therm. 20°. For a square 

 the rate is 2-22 and 2-29, average 2-25 cub. centims. per minute. The proportion of 

 oxygen was, in the first experiment 42-5, and in the second 41-66 per cent. 



