normal standard. It is impossible to apply to any individual patient the average data obtained by observations on healthy human beings, however numerous. If the abnormal measurements are to be of any value they must be compared with records of the same tests applied to normal parts of the patient himself.
Suppose, for example, that the sensibility of the right hand is affected; all the recorded measurements must be compared with those obtained from similar portions of the left hand. But if we attempt to work out on the normal hand the exact point at which any test can be appreciated, the patient's attention is exhausted before we come to the affected parts. In most cases it is impossible to work out a true sensory threshold even over normal areas of the body under the conditions of time and opportunity yielded by clinical medicine.
We have, therefore, adopted the following relative standard of comparison. A stimulus is selected which we know to be considerably above the normal threshold for the particular part of the body under examination. This can be appreciated without fail provided the patient possesses the usual intelligence and goodwill. On applying the same test to the affected parts, it may reveal more or less gross defects in sensibility; the strength of the stimulus is then gradually increased until either a threshold is revealed or no series of correct answers can be elicited by any stimulus applicable within the conditions of examination.
Let us take as an example of this method the use of the tactile hairs. We know from experience that contact with a hair exerting a force of 21 grm./mm.2 (5 grra./mm.) is above the threshold for the tips of the fingers in most persons. On the normal hand a series of such stimuli should lead to a perfect set of answers; this can be rapidly determined in any patient under examination, and we then proceed to test with the same stimulus equivalent parts on the affected side. If the answers are defective the strength of the stimulus is increased by selecting hairs of greater bending strain until each contact is appreciated. It may be that this condition is not reached until the test-hair exercises a pressure of 70 grm./mm.2; we then know that stimuli of this strength are necessary to evoke a constant series of answers over the affected parts, although the normal hand is sensitive to a hair of 21 grm./mm.2
We make no attempt to obtain an absolute threshold on the normal side; we establish a relative difference in sensibility, and so avoid the long and wearisome procedure necessary for a strict psycho-physical determination.
An exactly analogous procedure is adopted for measuring the appreciation of passive movements or the power of discriminating the two compass points. We choose a stimulus which is demonstrably super-normal in each individual patient. Then we determine by how many times its strength must be increased to evoke a series of correct answers over affected parts. Sometimes this is not possible, because the replies remain imperfect even with the higher degree of stimulation compatible with the special conditions demanded by
