36 T ranaaciionii. 



nitrogen was drawn over into the pipette J by lowering the head and trans- 

 ferred into burette, and analyzed with alkaline pyro-solution in a Hempel 

 pipette, to obtain the percentage of nitrogen present. 



From the weight of hydrate drawn in, and the change of pressure after 

 the hydrate had been let in, the quantities of aldehyde and oxygen at the 

 end of the experiment could be calculated. 



The volumes of the reaction-vessels used were between 90 and 100 cubic 

 centimetres. 



Importance of cleaning the Reaction-vessel. 



Mention is made by Dr. Ewan of the fact that it seems to be of some 

 importance to keep the apparatus as clean as possible. This was found 

 to be of the utmost importance. The chief diflficulty experienced with these 

 experiments was that the aldehyde tended to decrease faster than it should, 

 according to the equation 



2C2H,0 + = 2C,H,02, 



this being due, no doubt, to polymerization taking place. Aldehyde poly- 

 merizes readily in the presence of mineral acids, bases, and many salts, 

 especially sodium-acetate, so that it is of the utmost importance that these 

 should be absent. No reliable correction can be made for polymerization, 

 and any experiment in which it was marked had to be rejected as practically 

 worthless. At the same time, the presence of the polymeric form seemed 

 to have a distinct retarding effect on the action, as will be shown later. For 

 this reason, and to make the conditions strictly comparable, the reaction- 

 vessel and the whole of the capillary tubing and vessel containing the 

 aldehyde were invariably left to stand at least twelve hours, filled with a 

 solution of potassium-permanganate and hydrochloric acid. The manganese- 

 dioxide was removed by oxalic acid, and everything washed out with dis- 

 tilled water with the utmost care, and dried in a current of hot air. 



Correction for Aldehyde dissolved in the Acid formed. 



By means of a factor ¥, calculated from the excessive decrease of 

 aldehyde as determined at the end of the experiment, on the assumption 

 that this decrease is due to the aldehyde being dissolved in acetic acid 

 according to Henry's law. Dr. Ewan has attempted to make correction for 

 this error. The value of this correction seems to me to be doubtful. In 

 the first place, the value of l' for any one temperature should be constant, 

 whereas the value of ¥ varies from 0-00103 to 0-002873 — that is, it is in one 

 case almost three times as great as in another — while the temperature has 

 only varied from 20-2° C. to 20-6° C. Further, this value of ¥ at 20-8° C. is 

 0-001767, or almost twice as great as at 20-6° C. This in itself is sufficient 

 to show that Dr. Ewan's constant is a very uncertain factor. There can 

 be little doubt that where the value of ¥ is at all great some of the decrease 

 in the aldehyde has been due to polymerization. 



Further, it is extremely unlikely that when such a vapour as that of 

 aldehyde, so near its condensing-point, had once dissolved in the acetic 

 acid under the fairly large pressures at the beginning of the experiment 

 it would vaporize again as the pressure decreased. 



In any case, in most of the experiments, the error due to the dissolving 

 of aldehyde by acetic acid is small, because the quantity of acid formed 

 is small, and the surface exposed is also small, as the acetic acid runs 

 down into the drawn -off part of the pipette. In those cases in which 



