402 MEMOIRS OF THE NATIOifTAL ACADEMY OF SCIENCES. 



to the ordiuary criteria which have been proposed for rejecting observations witli large residuals,, 

 two methods adjusted to the present work. One of these was for the subject, after he had made a 

 reaction, to judge whether or not it was normal, and to assign its weight. We found, however,, 

 that directing the attention to the reaction interfered with its reflex character, and that it was. 

 difficult to assign a weight. We did not, therefore, continue this plan. We also applied a method 

 for rejecting the more discordant observations. In each series we made thirteen reactions, aud 

 rejected the time which departed most from the mean, then the time which departed most from 

 the mean of the remaining twelve reactions, and finally the third most discordant time. We thus, 

 had the mean of the ten most accordant reactions, which would represent a compromise between 

 the median and the mean. We did not, however, continue this method, owing to the considerable 

 calculation involved and to the fact tiiat the corrected mean departed very little from the mean of 

 all the observations. It is evident, however, that a result wonld be reached more quickly if some 

 objective method were adopted which would exclude observations which are not normal, and we 

 hope at some future time to consider the result of applying various criteria, to our actual experi- 

 ments. In this paper aud others previously published we have made over twenty-five thousand 

 separate measurements of the times of i>nysiological and mental processess, aud have consequently 

 ample material for studying methods for adjusting errors of observation. 



Part II. — Reactions on Electrical Stimuli. 



An electric shock can be applied conveniently to different parts of the body. We used a 

 Da Bois-Reymond induction coil (cf. rig. o) and the shock following breaking of the primary circuits. 

 By means of a double key (flg. i] ou the gravity chronometer (which was at first closed by the- 

 falling screen and later by hand) the primary circuit was broken and the current controlling the 

 chronoscope simultaneously closed. We had supposed, from theoretical considerations, that the- 

 induced current might occur at a not inconsiderable interval after the breaking of the primary 

 circuit, and that this would explain the fact that the times were longer for electrical stimuli than 

 for touches. By the kindness of Dr. Scripture we tested this ou a chronograph, but could discover- 

 no appreciable interval between the breaking of the primary circuit aud the spark from th& 

 induced current. 



Various electrodes wei^e used to apply the shock to the skin. We used electrodes 5 cm. apart 

 and electrodes (platinum surfaces 10 mm. in diameter) which could be adjusted on opposite sides, 

 of the limb. The method we found best was to apply one electrode (usually 10 mm. in diameter) 

 to the skin at the point we wished to stimulate while the other was conducted to.a pail of saturated 

 salt water iu which the left foot aud leg were placed. Electrodes were applied to several parts of 

 the body, and the current could be switched to any electrode (flg. 5). The stimulus was given ten 

 times in succession at the same point and then switched to another point immediately and without 

 shifting the electrodes. The shock was usually given on the left-hand side of the body, the 

 reaction being made with the right hand or foot. 



The sensory effects of electrical stimulation of the skin have not been properly investigated. 

 While works on physiology and psychology (e. g. Hernumn, Foster, Wundt) discuss in detail the 

 effects of electrical stimulation on the organs of sight, hearing, taste, and smell, they are curiously 

 silent as to the sensory effects of electrically stimulating the skin. These effects are varied and 

 interesting, and deserve a more careful investigation than we were able to make. In the case of a 

 galvanic current from twenty-eight gravity cells in pairs the sensory effects on C were as follows: 

 When the current passed through the body to the foot iu saline solution no sensation was felt in 

 the foot. When the positire pole was applied to the dry skin of the arm no sensation occurred on 

 making, breaking, or with continuous current. When the skin was moistened with saline solution a 

 slight i)rickling was felt on closing, not perceptible afterwards nor ou breaking. When the iieciafire 

 pole was applied to the moisteued skin of the arm there was' a slight sensation on closing, then 

 after a brief latent period a sensation as of piercing and boring followed, which became very 

 painful. The stimulation caused tetanus of the muscle and left blisters on the skin. When the 

 Ijositive pole of the same current was applied to the dry skin of the outside of the upper lip there 

 was on closing a slight ]n-ickling sensation, a slight flash of light, and a slight metallic taste. 

 When the lip was moistened these were all stronger, the prickling and taste continued while the- 



