88 SCIENCE PROGRESS 



axial ratio of which, however, is not the ratio of compression 

 to transverse extension, as it would be if the gel remained 

 uniform throughout. 



Solutions which are reduced by the gelatin itself, like silver 

 nitrate or potassium permanganate, proved unsuitable for 

 similar reasons, although they also produced sharp boundaries. 

 The gel in the reaction zone is incidentally altered to such an 

 extent that its modulus differs widely from that of the surround- 

 ing material, so that the results are again spurious. 



After numerous experiments, detailed enumeration of which 

 is unnecessary, the following arrangement proved completely 

 satisfactory : a 0*5 to i-o per cent, solution of potassium chro- 

 mate diffuses into a gel containing o*i per cent, of lead nitrate 

 or acetate. The lead chromate formed by the reaction is 

 highly disperse, does not precipitate in Liesegang rings, and 

 gives an exceedingly sharp boundary. Careful examination 

 failed to show any sweUing under the chromate solution. Puri- 

 fication of the gelatin — which lowers its elastic modulus to an 

 undesirable extent — is unnecessary, as any lead salts possibly 

 formed, such as carbonate or sulphate, are more soluble than 

 the lead chromate and accordingly react with the potassium 

 chromate solution. With the concentrations stated above 

 diffusion zones which can be measured accurately are obtained 

 within four to eight hours, according to the gelatin content of 

 the gel ; if necessary, the experiments can be extended for 

 considerably longer periods, as the lead salts appear to have 

 a marked antiseptic effect, so that specimens can be kept at 

 room temperature in an atmosphere saturated with water 

 vapour for five to eight days without showing mould or signs 

 of putrefaction. 



The second difficulty was that of letting diffusion take place 

 through an opening of strictly defined size. In the initial 

 experiments, masks of lead, tin, silver, and rubber foil were 

 employed, into which a circular hole of 2-3 mm. diameter had 

 been punched. These masks were squeezed on the gel surface 

 so as to secure good contact, and the chromate solution was 

 then placed on the opening with a small pipette. With these 

 opaque masks it was impossible from the outset to be sure 

 that perfect contact round the edge of the opening had been 

 secured ; even if this (as far as could be judged from examination 

 of the specimen at the end of the experiment) had been the case 

 at first, the solution frequently crept under the mask. Per- 

 fectly satisfactory results were finally obtained with glass 

 masks consisting of halves (i in. x i| in.) of thin microscope 

 slides, with a drilled hole i -i mm. diameter in the centre. With 

 these transparent masks the absence of air bubbles immediately 

 showed whether complete contact with the gel surface had been 



