[LM111488], Test Report, Arthur Edwin Kennelly, July 1889


View document with UniversalViewer   → View document on Archive.org  → Re-use this digital object via a IIIF manifest


[LM111488], Test Report, Arthur Edwin Kennelly, July 1889

Editor's Notes

The Law of Probability of Error as applied to the observed electrical resistance of the human body. ## It is generally admitted that all social phenomena are in the aggregate controlled by the general law of probability of error and statistics it is claimed shew that the propositions, weight, intelligence, and in fact all measurable properties of animated beings, at any given time and place center about some mean value the divergences from which follow the law of probability of error in observations. This leads to the inference that Nature endows all organisms with slight variability of type in a high degree, any one quality being the resultant of a great number of factors, so that the general mean is reached through the summation of all possible variations of these factors, as many variations ultimately positive as negative. The law has probably not been stated for the measured electrical resistance of the humans body, but the folowing observations may fairly be said to indicate its application.####Two hundred and twenty men and boys were measured at Mr T.A. Edison's Laboratory for the resistance they offered between the hands with liquid contact. The measurements were all made on the same day and as nearly as possible under the same conditions. Two glass jars 15 cms (6") --- in diameter and 20 cms (8") deeo containing an aqueous solution of caustic potash (density 1.10)-- were connected by immersed copper 15 cms (6") x 39 cms (4 1/2") with a Wheataton bridges. The reason for adopting caustic potash solution was that in a previous test with thirty nine men their mean resistance ---- the hands immersed in this solution was found to be 30% less than when square zinc sulphate solution of density 1.13 and zinc electrodes were used, all other conditions remaining unaltered####[note" something scribbled on the side]####[page 225]####The testing battery had a low resistance and an e.m.f. of 8 volts. The bridge ratio was 1000/1000 and with the dead beat galvanometer employed reading could easily be taken to one ohm in 2000.####The measurements were made in such a manner as to eliminate almost entirely the effects of polanisation. And these circumstances it was found that between the limits of the feeblest current by which an observation could be obtained, and the charged current that could be conveniently supported by the men (about ten milliamperes) the measured resistance were practically independent of the current strength and consequently of the battery power employed. The resistance however decreased with the surface area of immersed skin, and also with the period of immersion, in many instances falling about 100 ohms in this first thirty seconds. TO obtain comparable results therefore the hands of the person tested were plunged pointing vertically downwards into the hars until the tips of the middle fingers rested on the bottom, and this measurement was made as closely as possible thirty seconds after immersions, The ---- -- The readings were carried to the ----- ten ohms, units being negligible even if the resistance had been steady enough to render them distinct.####The variations in these measurements would include not only differences in the resistance of arms and ---- but also the variations of thickness ------- the epidermis area od hand surface, as well as error of observation and time. All three disturbing conditions however may be regarded as belonging to the class of observational error, and simply tend to make the measurement of the resistance within the body less precise. The age, height, and weight of the subject were recorded with each observation. No dependence of the resistance upon these ----- could be noticed, but the boys had apparently a higher resistance and stoutly built####[page 226]####men a low resistance. Thus the four boys of ages under 16 made an average of 12.18 while the mine case of men whose weight exceeded 175 lbs made an average of 770-####The average of all the 220 measurement was just 1000-, and the sums of the divergence from this mean value taken without regard to sign was 27,920.####From these data we obtain by the application of the usual approximate formulas of probability the following tables in which the actually observed and calculated divergences are compared.####[chart]####The agreement is seen to be very fair except that there were more large discrepancies of over 500 or than theory would account for.####The highest observed resistance was 1970- and the lowest 660.####Under the conditions of the test therefore it was an --- chance by the --- of probability that any means resistance would be within 108- of 1000 i.e. would lie between 892 and 1108, As a fact 112 out of the 820 observations fell within those limits, so that the final result of all the measurements was 1000 - 100.####[name mentions: Arthur Kennelly, Francis Upton, Roberts Brevoort Electric Co Ltd, Thos Edison, W.J Jenks, W.J. Hammer, F.S. hastings, Sir John Pender, Mr Lundy, C.J Goodrich, Professor Eggleston, Dr C MacDonald, Professor Chandler, Mr Graves, C.G. Ward]





Folder/Volume ID


Microfilm ID


Document ID



Thomas A. Edison Papers, School of Arts and Sciences, Rutgers University
Download CSV | JSON