MancJicsti'i- Mcinoirs, Vol. h. (191 1), No, 10- 5 



4 14 grms. of orthoboric acid yield 292 grms. of 

 metei acid and 264 grms. of p)-ro acid. In this case the 

 loss through volatiHsatiou is much greater, the final 

 product weighing less than the tlieoretical amount for 

 boric anh}'dride. It would, therefore, seem that the 

 amount volatilised depends upon the rate of heating. 

 These three experiments, only calculated to show the 

 percentage of water present in the compound from time 

 to time, are shown graphically in Figs, i and 2, from 

 which it will be seen that only one break exists in the 

 deh)-dration curve. It will be noticed in Figs, i and 2 

 that the break in the curve occurs with a lower percentage 

 of water as the temperature at which the experiment was 

 conducted was raised. This is due to volatilisation of the 

 ortho acid in the water vapour. The difference between the 

 percentage of water present in the compoutid when the 

 break takes place and the theoretical value for metaboric 

 acid (205 per cent.) gives a rough measure of the volatility 

 of the ortho variety. 



These results confirm those of other experimenters as 

 to the marked change in the rate at whicli the ortho acid 

 loses water, and probabl\' show the definite existence of 

 metaboric acid, but there is no indication of any other 

 compound. 



I'he second series of experiments consisted in the 

 determination of the melting points of mixtures of ortho- 

 boric acid and boric anhydride. Intimate mixtures of the 

 required composition of the two substances in a fine 

 powder were made, and then introduced into stout 

 capillary glass tubes, which, as soon as they were filled 

 were sealed as close as possible to the top of the mixture. 

 The melting point was then determined in a bath of 

 sulphuric acid. It was necessary to use sealed tubes as 



