this is often difficult, and care must be taken to guard against this source of error before deciding that sensibility to pain is disturbed solely on the ground that pricks are described as "less sharp" than over normal parts.
If the loss is slight, it becomes necessary to determine the threshold for pain. We have consequently employed, as a rule, a simple form of spring algesimeter (fig. 2). It consists of a metal tube (A) about 15 cm. in length, closed at one end and containing at the other a piece of vulcanite (B), flattened at its projecting end and perforated to allow the projection of a needle (C). The tube contains a fine steel rod, to one end of which this needle is attached. A fine spiral spring is fixed to the blunt end of this rod, and the other end of the spring is inserted into a small bar (F) which projects into the tube through a slit (a ... b) in one side of it, and is carried on a collar (E) that runs on the outer side of the tube. The spring is so arranged that it exerts no traction on the needle when the collar is at the highest point of the slit, and if the instrument is then applied vertically the point of the needle bears its own weight only. If, however, the collar is slid down towards the point of the needle tension is put on the spring and exerts a corresponding pressure on the needle. By measuring this on a balance the instrument can be graduated according to the pressure in grammes exerted on the needle, when the collar stands at different points of the scale. An instrument graded between 2 grm. and 10 grm. is sufficient for ordinary clinical purposes. When it is used horizontally, or with the point upwards, these values vary according to the weight of the needle, but this variation can be easily calculated. When, however, as in our work, it is sought to obtain a relative or comparative rather than an absolute threshold, this is unessential, provided the instrument be applied at the same angle to the corresponding points of the two sides of the body.
It has always been recognised that it is difficult to obtain an accurate threshold for painful prick; for if a pin be applied with the same moderate pressure twice in succession to the same part, one contact may be appreciated as pain and the other as touch, depending largely on whether a pain-spot is directly stimulated or not. We consequently apply the algesimeter a certain number of times in close succession to the part to be examined, asking the patient to say whether he appreciates a prick or merely a touch, and take the reply for this series of stimulations instead of for each individual one.
We have found the interrupted current an unsatisfactory means of measuring sensibility to pain, and have not used it systematically in those researches.
(b) Pressure pain.—Whenever pressure is appreciated after division of a
