In these results, air and ether vapour were drawn by a pump through the material to imitate the inspiration, but no attempt seems to have been made to imitate expiration. The effect upon the material used of moisture condensed from the expired air is not taken into account in these experiments. This is a serious hiatus in the argument, particularly as regards lint. This material in actual use rapidly becomes quite sodden, and ether will not vaporise from it properly. In spite of this fault, these observations may probably be taken as being reasonably accurate. With them may be compared the figures of Karl Connell, who, working with quite accurate methods, estimated the percentages of ether necessary to induce and maintain anæsthesia:-- Period of Anæsthesia. Percentage. First 5 minutes (_i.e._ induction) 18 Next 25 „ 14 Next 30 „ 12 Next 60 „ 12·8 “Bad” subjects on the average required an extra 4 per cent. during first half hour, feeble patients required 2 per cent. less. (_Journal_ of the American Medical Association, 22nd March 1913). APPENDIX III. THE ACTION OF ANÆSTHETICS UPON THE BLOOD. _The Blood Gases in Anæsthesia._ Buckmaster and Gardner (_Journal of Physiology_, vol. xli., p. 246), analysed the blood gases in various stages of chloroform anæsthesia, and some of their results are shown below in tabular form. They show a very definite reduction in the oxygen content of the blood. So far as one gathers from the text of the paper, the animals were not subjected to any considerable trauma during the progress of the anæsthesia, so that the figures arrived at with regard to the CO_{2} content do not bear upon the Acapnia question. If there was no trauma, there would be no deep breathing, and a reduction of CO_{2} could not be expected.