382 HEART DISEASES [VOL.VI sen es as a gauge to measure the amount of oxygen the patient is getting. The gas should pass through the water at a rate of at least five bubbles per second. Blood-letting to the extent of 10 to 20 ounces may prove a life-saving measure in acute heart failure. The special indications for employing this treatment are venous engorgement with cyanosis and acute pul- monary oedema. (4)—Surgical Measures Certain surgical measures which have been employed in the treatment of heart failure merit consideration, though they must still be regarded Tfyruid- as being in the experimental stage. Complete thyroidectomy by lowering ectomy fae basal metabolism reduces the heart's liabilities, and might therefore be indicated on theoretical grounds. In congestive heart failure the results of this operation ha\e not corne up to expectation, but in angina S\mpath- they have been more promising. Operations such as sympathectomy tctomy serve to reiieve pain? but whether these procedures are justified appears doubtful, for pain in angina is a danger signal which warns the patient that lie is overtaxing his heart; if therefore the pain is relieved without eradicating the cause of the disease, an important safeguard may be- renioved. (5)—Summary In Fig. 57 I ha\e attempted to summarize diagrammatically the therapeutic possibilities in the treatment of heart failure of the ischaemic type. A similar diagram could be devised for the congestive type. In the normal healthy subject (I), the cardiac reserve (A-E) far exceeds the basal resting requirements of the body (E-F). In the patient with angina (II), on the other hand, the cardiac reserve (C-D) is greatly reduced. The reduction is in part attributable to impaired efficiency of the heart (A-C). This depends on various factors, of which the most important is insufficiency of the coronary circulation (A-B). Structural changes in the coronary arteries and in the myocardium are permanent and therefore beyond the scope of therapeutic measures. But, as has been pointed out above, the nervous factor is of great importance in determining not only the sensitivity to pain but also the liability to attacks. This can be, in part at least, controlled by cerebral depressants. Similarly, toxic factors affecting the myocardium, such as tobacco and focal sepsis, may be eliminated. By these means efficiency of the damaged heart can be increased from D-C to D-C'. It may be hoped, by suitable treatment, to lighten to some extent the load imposed upon the heart by such factors as obesity and high blood-pressure, thereby increasing the cardiac reserve from C-D to C-D'. That marks the limit of possible therapeutic achievement. When it has been attained, the patient must be content to adapt his mode of life to his reduced exercise tolerance C'-D'.