220 Frederick Guthrie on Salt Solutions 



solid colloid walls are broken, the liquid contents of the in- 

 dividual cells become continuous, so that all hydrostatic laws 

 hold good. Let the heated liquid cool, and similar cells are 

 restored ; but on repeated heating, the cells are burst in such 

 divers places that they are not to be restored. Such re- 

 storation is also hindered and prevented on stirring during 

 cooling. 



I gladly leave these speculations for the facts which gave 

 them birth. Through the kindness of Mr. W. C. Roberts, 

 F.R.S., there were placed at my disposal certain of the inor- 

 ganic colloids prepared by the late Mr. T. Graham, the Master 

 of the Mint. Some of these are spontaneously undergoing so 

 interesting a change that I have been unwilling to disturb 

 them. Others, notably a solution of silicic acid and of ses- 

 quioxide of iron in water, I find contain each under 1 per 

 cent, of oxide. I prepared, therefore, fresh solutions, follow- 

 ing the directions given by Mr. Graham, and meanwhile ex- 

 amined certain organic colloids. 



§ 162. Gum arabic— -If a 1-, 5-, 10-, 20-, 25- or 30-per-cent. 

 aqueous solution of gum arabic be cooled, ice begins to be 

 formed in all cases at 0° C. exactly. Even a 35-per-cent. 

 solution, which is very viscid, begins to give up ice at 0°, and 

 is quite solid at — o, 5. A 45-per-cent. solution, which is 

 semisolid, shows a temperature of — 1° at the beginning of 

 sensible ice-formation, and is certainly solid at — 1°*2. This 

 very small degree below 0° C. in so viscid a substance is cer- 

 tainly not more than may be attributed to the overcooling of 

 the ice already formed. I can find no trace of a cryohydrate; 

 and that such does not exist is supported by the fact that, in 

 spite of the great solubility of gum in water, when powdered 

 gum and ice are mixed together the resulting temperature is 

 0° exactly. Gum arabic is therefore powerless as a cryogen. 

 It is true that this might be attributed to the gum acting like 

 a certain quantity of anhydrous chloride of calcium, which 

 when mixed with ice may compensate, by the liberation of 

 heat of combination, for the heat absorbed in the joint lique- 

 faction. But that such balancing does not ensue is shown by 

 the fact that a 45-per-cent. solution of gum also gives 0° C. 

 with ice, while a solution, however weak, of chloride of cal- 

 cium will absorb heat when liquefying ice by contact. The 

 same is also proved by the observation that when gum is 

 mixed with water the alteration in temperature, though cer- 

 tainly in the direction of heat-liberation, is very inconsiderable. 

 In a preliminary experiment I found that powdered gum and 

 water mixed in indefinite proportions gave a rise of o, 75 C. 

 As the result of a number of more exact experiments, it ap- 



