With reasonable skill and experience and the utmost care, it is, however, perfectly possible to prolong nitrous oxide anæsthesia for periods of five, ten, or even fifteen minutes in the average healthy patient. As soon as the signs of full anæsthesia appear, the valve tap is pushed back to “air” for the space of one inspiration and one expiration, and then at once pushed back to “no valves.” By this manœuvre, one inspiration of air is permitted to the patient, whose colour at once shows amelioration, or at any rate no further progression of cyanosis. The admission of air is repeated every third, fourth, or at most fifth respiration. After the first minute or so of this cycle of events, it is obvious that the contents of the bag will be composed of a mixture of nitrous oxide, air, and CO_{2} in proportions quite impossible to calculate. It is therefore best to push the indicator to “valves,” and allow the bag to be emptied by the suction of the patient’s inspirations. The cylinder head is then opened by the turning of the foot-key, and the bag filled again with gas. The cycle of “air” and “no valves” is then begun again for another minute or so. It must be understood that by this process, it is not to be expected that an ideal anæsthesia can be produced. Some movement of the patient will not improbably take place when sensitive structures are cut or handled by the surgeon, and at no time will the muscles be entirely relaxed. Such an anæsthesia is therefore only suitable for a limited class of case, but does admirably for, say, opening an abscess, exploring its interior, and removing from it a sequestrum or an easily found foreign body. During his service in Macedonia the author had not at his disposal any of the appliances later to be described under the heading of nitrous oxide and oxygen, and found “gas and air” a most useful form of anæsthesia for the requirements of military surgery as seen in a Base Hospital, under active service conditions. Contra-indications to the Use of Pure Nitrous Oxide Gas. In the healthy subject, there is no safer anæsthetic than nitrous oxide when administered properly and limited to its proper province. From the account given of the physiological action of the gas it will, however, be obvious to the student that in a limited class of case its use is not permissible. Such cases fall into two categories.