In cases in which a special _cooling of neck and back of head_ had existed before death, the _autopsy_ showed _a marked brain oedema_, a tight filling of the general brain cavity [Hirngefaesse] blood in the spinal fluid as well as blood in the Michaelisrhomboid. The heart findings warrant our taking a certain attitude toward _the question of rescue collapse_. Death occurred relatively quickly after removal from the water, which may be compared with rescue. The longest interval involved was 14 minutes. It is to be noted, however, in the first place, that almost certainly a much larger number of deaths would have been observed if an active heat therapy had not almost regularly been coupled directly with the completion of the experiment; in the second place, that in such cases there would have been very much longer intervals. We have already called attention repeatedly to the after-cooling following the experiment. In every case where this had proceeded to a certain point, countermeasures were taken, since the experiments were never planned to end in death. One may well imagine, however, that in mass catastrophes, in which almost exclusively rescue collapse has heretofore been described, the therapeutic measures were confined to an undressing and drying off of the rescued together with a subsequent wrapping in covers. Under these conditions after-drops of great magnitude and long duration were to be expected. In the course of this delayed fall in temperature, a heart-death might occur as in our experiments. _We should like to emphasize that the irregularity per se is not to be regarded in our experiments as a symptom of danger to life any more than in the clinic, but rather as a sign of direct heart damage, which increases continuously with further falling off of temperature, until finally the heart fails. If the temperature drop is arrested, the slow form of irregularity passes over into a rapid form._ This transition is a favorable sign for survival; for this irregularity virtually always passes over of itself after a time averaging 90 minutes into normal heart activity. It continues therefore for a long time after the body temperature has already risen markedly. A danger to the circulatory system could not be demonstrated at this stage. In three cases the return of the heart action to normal occurred in spite of simultaneous energetic physical work. With the demonstration that cold-death of man is primarily a heart-death, the essential points for therapy are also cleared up. The _cause of the severe heart_ damage is another question. Since our studies were primarily aimed at the development of practical methods of treatment, we will not go very far into the theoretical concepts which may be developed in this connection. Still, several hints may be drawn from the blood studies: 1. The great increase of the _viscosity_ causes an _increased loading_ upon the heart.