Similar treatment of corpuscles five days subsequently gave no indication of the presence of any methemoglobin spectroscopically. From the serum and from the laked corpuscles of number 784 no trace of methemoglobin was found when the blood was examined a few hours after removal at autopsy. _Oxygen Capacity._—The total oxygen capacity was determined by the Van Slyke method (by Dr. Rohde and Mrs. Macklin). At this stage the more pronounced type of influenza had subsided, but in early convalescence the capacity was within normal ranges. Other studies using different technique gave concordant results, but there were indications that oxygen was more slowly absorbed than normally. _Oxygen Content of Venous Blood_ measured by the Van Slyke method (by Dr. Rohde and Mrs. Macklin) on the same bloods examined for total oxygen capacity seemed to indicate a mild deficiency as compared to normal bloods. _Gases, Kinds, Quantity and Rate Yielded to Vacuum._—In general it may be said that quantitative differences observed are not considered fundamental, but that the studies indicate abnormal slowness in oxygen absorption. _Gases, Quantity and Rate of Absorption on Exposure to Air After Extraction by Pump._—The results emphasize slowness of oxygen absorption as compared to normal blood. The material to be examined was exhausted for three minutes in the receiver of the Van Slyke apparatus. One c.cm. was then transferred, with as little exposure to air as possible, to a small empty bottle, which was then closed and placed in communication with a calibrated, horizontal tube, containing a segment of alcohol, which served the dual purpose of a seal and an air volume change indicator. (See Fig. 1.) The apparatus was made in duplicate and mounted on a common base, so that simultaneous readings on different samples could be made. After establishing the zero position of the alcohol segment, the base on which the bottles were mounted was vigorously shaken in a uniform manner. Ten seconds after the period of shaking, the volume readings were taken. Successive periods of shaking and reading were conducted at 30-second intervals, until the test was completed. Actual volume changes were then calculated, tabulated and plotted. The greater confidence is placed on the results obtained by observing the color of the blood, as described below; but since then the method has been checked up and the results indicate that the findings were of sufficient accuracy to warrant their inclusion in this report.[1] Footnote 1: Studies along this line are being made with improved apparatus, the results of which, together with the description of the apparatus, will be published elsewhere. (See Am. Gr. Physiol., 1920, li, 195.) [Illustration: FIG. 1.]