But, evidently, we need to clarify our notions as we rearrange our facts. Somatic death is, after all, a most complex process. It may be shortened by instant and complete incineration, but scarcely in any other way. Even dynamite would scarcely simplify the problem. As to conscious death, that is _probably_ (though not certainly) a matter of seconds only or possibly fractions of a second. While we have no accurate appreciation of what constitutes consciousness, nor even just where it resides, the central nervous system appears to be its most probable seat. But conscious death may occur almost instantly without injury to this system, as when a bullet passes through the thorax and the heart, without injuring the spine. But what is it that suddenly checks all concerted and interdependent activity? Or does something or some controlling agency suddenly leave the body? A recent theory, having features to commend it, is to the effect that life is a property or a feature of the ultimate _corpuscles_ which compose the atom. Since these corpuscles bear to their containing atom a relative size comparable to that of the tiniest visible insect winging its way in a large church edifice, the intricacies of this particular theory readily appear. But it does seem as though among ourselves life has much to do with the hitherto neglected and despised nitrogen atom or molecule, since life inheres _par excellence_ in nitrogen compounds. Moreover, _vitality is conspicuously a feature of those chemical elements which have the lowest atomic weight_, while at the other end of the table of atomic weights stands radium, of whose destructive emanations I have already spoken. Another phase of the general subject of thanatology was suggested especially by Osier, who a few years ago called attention to the fact that but few, if any patients really die of the disease from which they have been suffering. This is not a paradox, and needs only reason and observation to confirm it. His statement was a preliminary to the consideration of terminal infections and toxemias, which of itself would be sufficient to erect thanatology into a dignified special study. Take, for instance, a patient who has long suffered from diabetes. The end is characterized by coma, i. e., an evidence of profound toxemia, and is in large measure due to acetonemia. A patient with chronic Bright's disease dies of uremic poisoning, or one with pneumonia dies of genuine heart-failure. The terminal stage of cancer is, again, toxemia of one kind or another, according as it has interfered with digestion, with respiration, or some other vital function, or has broken down, thus saturating the patient with septic products.