THE PULSE-RATE IN VERTEBRATE ANIMALS 73 



There may be other movements, but these first quick ones, 

 each the precursor of a ventricular systole, are easiest to count 

 when recorded. To record the movements the image of the 

 tip of the tube is magnified some 300 times and the boundary 

 between mercury and acid is photographed on a moving plate 

 on which is simultaneously projected the shadow of one end 

 of a tuning fork vibrating at a known rate, so that the speed of 

 the plate may be gauged. Fig. 3 is two-seconds' worth of a 

 record taken with a goldfinch arranged in the way shown 

 diagrammatically in fig. 2. The tracing of a tuning fork vibrating 

 100 times a second is seen above, and a thick and a thin hori- 

 zontal line which do not here concern us; the white below is 

 the acid and the black the mercury. The record reads from 



Jovk 



Fig. 3. — Electro-cardiogram of a goldfinch. 



right to left and it will be seen that the acid moved towards 

 the mercury at regular intervals. These can be counted ; in 

 this particular photograph 30^ of them occurred in the two 

 seconds, indicating that the heart was beating at the rate of 

 915 times per minute. 



Until this method was introduced the frequencies of beat 

 in small warm-blooded animals were not actually known. Their 

 order had, however, already been inferred by Dr. Haldane 

 from the known quick rate of consumption of oxygen. The 

 method he introduced some fourteen years ago of detecting 

 the presence of carbon-monoxide in mines, which has been the 

 means of averting many disasters, depends essentially upon the 

 fact that the more rapid the circulation is through the lungs, 

 the more quickly is an animal affected by poisonous gases 

 absorbed from the atmosphere and the more quickly does it 

 recover in air free from such gases. Since carbon-monoxide, 

 which is far more dangerous to life than any of the other gases 



