30 Frederick Guthrie on Salt-Solutions 



departures, and is now that of saturation at the new tempera- 

 ture in the two cases. The surface of demarcation shifts 

 towards the centre of the mass. Similar effects follow further 

 rises of temperature. Let now the lube cool. The former 

 condition is reestablished. On further cooling, the surface of 

 demarcation shifts — rises in this case as the two liquids mix 

 more with one another. At 21°*3 perfect solution ensues, 

 and this is well marked within a tenth of a degree. Just 

 above this temperature the mass is opalescent, just beneath it 

 it is perfectly limpid. When at this critical temperature, the 

 warmth of the hand only momentarily applied makes a wide 

 test-tube of the limpid mixture quite milky. Nor can such 

 a test-tube be passed so nimbly through a flame as not 

 to exhibit the same change. Following the temperature 

 downwards, we find the solution remains clear below 0°, until 

 we reach — 2°, whereupon crystals of ice appear. These 

 increase with the falling temperature until — 3°*8 temperature 

 and 10*1 per cent, is reached. The cryohydrate now solidifies, 

 and no further percentage change is possible ; nor tempera- 

 ture change, until solidification is complete. 



A 40-per-cent. ratio only shows perfect mixture at 18°'4. 

 It remains a mixture until — -5°*1, whereupon subcryohydrate 

 begins to appear in the solid form. Its quantity goes on in- 

 creasing with lowering temperature until —20°* 6 is reached: 

 this marks an 80-per-cent. solution. At still lower tempera- 

 tures no doubt the liquid becomes more enriched. If the 

 anhydrous liquid is unsolidifiable by cold, then the ordinate at 

 100 per cent, is asymptotic; but if, as is more likely, the liquid 

 is solidifiable, the remainder of the water and of the ethyl- 

 amine will unite to form some hydrate even lower than the 

 subcryohydrate at some very low but constant temperature. 

 An 80-per-cent. ratio at — 30° is thick with solid subcryo- 

 hydrate. Perfect liquefaction takes place at — 20 o, 6. The 

 liquid remains clear until + 12° is reached. At and above 

 12° separation ensues, which increases with the increasing 

 temperature. The curve (fig. 1, " Triethylamine and Water 

 as liquids") is virtually a straight line between 15 per cent. 

 and 50 per cent. The deviations are probably errors of obser- 

 vation; and these errors of observation are certainly not due 

 to want of definition of the occurring and vanishing turbidity, 

 which is well marked within 0*025 of a degree. Nor are they 

 due to errors of original weighing or change during mani- 

 pulation ; for the very flatness of the curve shows that a great 

 change of composition has little effect upon the temperature 

 of saturation. If they are indeed errors, they arise probably 

 from the collecting together of the opalescing globules into 



