FAST AND SLOW RECORDS OF FATIGUE 
207 
three hundred and fifty responses in undergoing fatigue. Both 
curves show that up to about the one hundredth response, the 
amplitude in terms of height is not greatly affected and although 
the first few responses are accompanied by primary contracture 
subsequent relaxation is remarkably gradual as is shown in the 
slightly declined margin of B' and the slight overlapping of re- 
laxing lines in B near the base. One of the most valuable facts 
in comparison is brought out in connection with relative dura- 
tion of the different phases. From the curve B' one would be 
led to believe that successive responses following the one hund- 
redth were being executed in about the same interval of time as 
those immediately preceding. On comparison with B, however, 
such is found not to be the case. Successive contractions here 
are shown to be more and more slowly produced, and in addi- 
tion a very marked slowing of the successive relaxation phases. 
Although, of course, it is difficult to be exact because of com- 
parative differences in various muscles it seems reasonable to 
assert that comparatively speaking the successive contraction res- 
ponses vary from one another by intervals of .001 second in 
this range between the 100th to the 200th response, while the re- 
laxation phase certainly is retarded by as much as .01 second. 
Beyond this up to perhaps the 300th response, the contraction 
response time increases until it takes a total of at least 0.2 second, 
while the relaxation phase is retarded so that at the end of this 
interval it consumes at least .08 second. Secondary contracture 
can best be recorded by the slow drum method, as also can 
secondary relaxation. These two phases- develop in the record 
under consideration between the 100th and the 200th responses, 
and between the 200th and the 300th responses, respectively. In 
the fast drum record B and the comparable slow record B' these 
phases make themselves evident as a respective rise and fall of 
the curves near the stimulation take-off of contraction, and as 
can be seen, neither are clearly defined. On the other hand, 
latency is not evidenced by the slow drum. By the fast method 
under favorable conditions and care of mechanical adjustments 
latency can easily be calculated as ranging in the initial responses 
between .004 to .005 second, and gradually this interval increases 
as fatigue develops so that toward the end it may be as much as 
0.02 to 0.04 second, depending upon temperature, load and os- 
motic conditions. 
Curves G and G' represent another extreme type characterized 
by marked secondary contracture with equally marked secondary 
