500 ANNUAL. KEPORT SMITHSONIAN INSTITUTION, 1910. 



not yet the determining data; for, however small the demand for 

 oxygen, all animals when active must consmne more than when at 

 rest and those that are habitually active must have some means of 

 obtaining more. Moreover, since a large heart works more eco- 

 nomically than a small one, in that it spends less of its time in over- 

 coming inertia, it would for that reason also be favored when much 

 work has to be done. The range of variation in relative heart size 

 is fairly large in all species in which it has been determined in sev- 

 eral specimens, but more so in some species than in others. Thus in 

 four specimens of the golden oriole it varied between 1.8 and 2.6 per 

 cent, while in seven of the curlew sandpiper it varied only between 

 1.6 and 2.0 per cent. In man, according to the determinations made 

 by Bergmann from 36 people in whom death was accidental, the 

 variation may be from 0.43 to 0.75 per cent. Miiller (2a) dealing 

 not with the weight of the whole heart, but only with that of the 

 musculature, in percentage of body weight, in a large number of 

 individuals who had died of different diseases, shows by his tables 

 that in about 800 people dying between the ages of 30 and 60, this 

 varied from 0.26 to 0.89 per cent, and further that the percentage 

 weights of this musculature did not vary symmetrically about a 

 mean, but asymmetrically about a mode (i. e., the percentage weight 

 of the greatest number), and in such fashion that the mode (0.49 to 

 0.50 per cent) was nearer to the relatively small hearts than to the 

 large ones, suggesting that the heart in man is becoming relatively 

 smaller. The suggestion that man's ancestors were larger hearted 

 is perhaps supported by the fact that in infants the modal ratio is 

 about 0.6 per cent and even in children from 4 weeks to 3 years of age 

 it is further in the direction of the large heart than in the adult, 

 being about 0.53 per cent. But we have to be careful in drawing 

 such inferences from data which can not be determined in the living, 

 since we do not know in how far the heart ratio affects the death rate, 

 a point which Miiller, who interprets his tables in a way very dif- 

 ferent from that which is here suggested, seems to neglect. How- 

 ever this may be, we have ample evidence that in man as in other 

 mammals, in birds, and so far as we know also in the lower verte- 

 brates, the material is there to be selected from should it for any 

 reason become advantageous for a species to alter its heart ratio in 

 the future as it has probably done in the past. With regard to the 

 past, it seems probable that such variations were used as material for 

 selection before they became correlated with frequency of beat and 

 that it was with the size of the heart more or less already deter- 

 mined that this frequency, which is also known to be variable in indi- 

 viduals, began to be used when it began to be advantageous to be 

 independent of external temperature, owing perhaps to a change 

 from an equable to a variable climate. In the present state of our 



