44 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



ences between the first and second reflex in each pair of records indicate 

 the relative amount of refractoriness of the patellar reflex of the sub- 

 ject at that particular time after that particular interval. 



The task of reading the records is rather exacting, though relatively 

 simple. At the rate for which the kymograph was regulated, each 

 running millimeter of the record is equivalent to 10 a (0.01")- The 

 records were read with a lens through a glass plate which was divided 

 into millimeter cross-sections. The glass coordinate system was so 

 placed on the record that its horizontal ordinates were parallel with the 

 base-lines of the records. It was then carefully moved so that a main 

 ordinate cut the record line at the first indication of reflex muscle 

 thickening. The length of the abscissa from the ordinate which was 

 placed on the beginning of the reflex thickening to the beginning of the 

 stimulation curve can be read on the millimeter scale directly to lOcr 

 (0.01") and estimated with reasonable accuracy to la- (0.001"). 



RESULTS. 

 VARIABILITY OF THE PATELLAR REFLEX. 



In all the studies of the patellar reflex its variability has been one of 

 the most conspicuous features of the records. Normal individuals 

 differ widely in their susceptibility to the ordinary stimuli. Since 

 the patellar reflex is not essential to any known vital process, these 

 individual differences are not surprising. In addition to the variation 

 between individuals, the patellar reflex is subject to more or less gross 

 variation in the same individual at different times. Even with string- 

 galvanometer technique, Jolly 1 found the latent time to vary in one 

 individual so that the highest value was more than double the lowest 

 (11. 7 a and 24.4 a respectively). When, as in Jolly's measurements, 

 the currents of action are used as indicators, this variation must be 

 almost entirely central. It is proportionately more prominent as one 

 decreases the relative importance of the peripheral factors. For the 

 purposes of a science that seeks an invariant, these central variations 

 seem unfortunate. On the contrary, no fact may properly be regarded 

 as unfortunate in science. The variability of the knee-jerk empha- 

 sizes, in the case of the simplest possible neural arc, the contention 

 which appeared in our introduction, that biological invariants do not 

 exist except as statistical artifacts. 



Simple reflex in an intact vertebrate is after all a relative term. 

 There is no reflex arc so simple as to consist of an isolated chain of 

 neurons from receptors to muscle-fiber. There is no reflex so simple 

 that we can conceive of it as a transmission of energy from receptor to 

 reactor through a more or less resistant conductor. At no step, except, 

 perhaps, in conduction through the axones, does the process follow a 

 physical model. In no living organism can we ever assume that an 



Molly, Quart. Journ. exp. PhysioL, 1911, 4, p 07; British Med. Journ., 1910, 2, p. 1259. 



