90 SCIENCE PROGRESS 



before clamping the test piece : this precaution was found to 

 be necessary on examining test pieces in polarised light, when 

 they were seen to adhere strongly to the jaws in places and 

 to be deformed in a highly irregular manner. Greasing the 

 faces prevents this adhesion and allows the test piece to expand 

 regularly in the two directions at right angles to the axis of 

 the screw spindle. According to the gelatin concentration 

 the prisms were compressed to 24, 22, or 20 mm., the initial 

 dimension being 28 mm. in all instances. Two test pieces 

 clamped in this fashion, and one control piece left quite free 

 (all three cut from the same piece of gel), were placed in a metal 

 box lined with moist filter-paper and were then covered 

 with the masks and the chromate solution. The progress 

 of diffusion, i.e. the formation of lead chromate, is easily 

 observed through the glass mask ; when the zone of chromate 

 had attained a sufficient size, the solution was removed with 

 a capillary pipette and filter-paper, the mask detached and the 

 diameter of the diffusion zone measured to o-i mm. 



The experimental arrangements have been described in 

 detail which may seem exaggerated unless it be taken into 

 account that all the precautions were proved necessary by painful 

 experience and that the accuracy of the results depends on their 

 complete and rigid observance. 



The outcome of numerous experiments was somewhat 

 unexpected — at least by me. Repeated investigation of 6, 8, 

 10, and 12 per cent, gelatin gels showed that the diffusion zones 

 in the stressed prisms were perfectly circular in all cases : more- 

 over, the diameter of the diffusion zone was the same in a compressed 

 as in an uncompressed test piece of the same gel. In other words, 

 the gel remains isotropic for diffusion under stress and the 

 diffusion velocity is the same in the deformed and in the un- 

 stressed gel. 



One set of experiments out of a large number performed 

 will be sufficient to illustrate the point. The gel was : 6 gm. 

 of Cox's powdered gelatin in 100 c.c. of water with o-i gm. of 

 lead acetate. 



If the prisms are released at the end of the experiment 

 — the diameter of the diffusion zone having, of course, been 

 measured before release — they expand again, but do not, not- 

 withstanding the short duration of the stress, return to their 

 original dimensions. Owing to this expansion the circular 



