THE ELECTRICAL SPASM OF DEATH 193 



which discoloration begins cannot be detected with sufficient 

 precision. When we place the specimen in a thermal bath 

 under a rising temperature, the beginning of discoloration 

 after death is so slight as to be impossible of detection, and 

 by the time it becomes marked, the temperature has already 

 passed several degrees above the fatal point. I have found, 

 for example, that the colour of the milk-white style of Datura 

 alba has changed to brown by the time that the temperature 

 of the bath has risen to about 64 C. In the petals ot 

 Sesbania coccineum, again, a striking change of colour is 

 detected, under similar conditions. Rich crimson here turns 

 into pale blue at a temperature of about 67 C. The fila- 

 mentous corona of Passiflora quadrangularis^ finally, in which 

 the filaments are barred by purple rings, loses its colour 

 normally at about 68 C. In all these cases, the initiation of 

 the loss of colour must have been imperceptible. Hence, all 

 that can be determined from such experiments is that the 

 death-changes must have commenced at some temperature 

 lower than 64 to 68 C. 



Before proceeding further, it is necessary to obtain a clear 

 idea of what is meant by the death-point. In animals, an 

 early symptom of death consists in the setting in of rigor 

 mortis. But this does not synchronise throughout the 

 body, certain parts of the organism undergoing the death- 

 change earlier than others. Thus the only definition of the 

 death-point which can be made at all precise is that which 

 regards it as the point of initiation of some unmistakable 

 sign of death. I shall next proceed to describe several 

 death-symptoms and the modes by which they may be de- 

 tected with certainty. 



With regard to such detection, I have pointed out else- 

 where that, theoretically, it should be possible to make such 

 a determination by watching the waning of some effect 

 characteristic of the living condition, the death-point being 

 known by its cessation at a given moment. Such a test, as 

 we shall presently see, is afforded by the electrical responses. 

 The ideally perfect method, however, would be by the 



O 



