14: 
A Stiidji in Criminal Anthropometrij 
of temperature, and -183 respirations per minute for every 10 beats per minute 
increase of the pulse. 
A further interesting point is that, if we could possibly suppose these relations 
for criminals, whose disorders, if any, do not enforce hospital treatment, to extend 
beyond the range of values observed, then the pulse would cease at a body 
temperature T,„ = 87°-05* and the respiration would cease at a body temperature 
2",„ = 79°-ll. We have no right, however, to extend our linear regressions to 
Diagram IV. Regression Lines of Temperature on Pulse for Feeble- and Normal-Minded. 
these extreme cases ; all we can say is that they suggest cessation of pulse before 
that of respiration, and raise the problem of whether the temperature at death, 
T„, , is or is not on the average as much below normal temperature as is indicated 
by the above vague figures. 
In Diagram III the regression line of temperature on pulse is drawn to test the 
general accuracy of linearity in describing this relationship. It will be seen that 
graphically the accordance is good, the group 45-48 of pulse-rate (see Table XXXIX) 
has only two entries and the deviation is non-significant. Algebraically for 
* Deduced from normal-minded criminals only T„[ = 80°-90. 
