The instrument commonly known as "improved form of Smedley's dynamometer" was used in this test.1 The handles of the instrument were adjusted to a span of 5.5 cm. and were kept constant at this distance with all the subjects. The subjects as a group and individually were given full instructions in the use of the dynamometer. They stood erect and free from other support; the instrument was held down at the side of the thigh, not against the leg, and dial outward.2 Five trials were made with each hand alternately. There was an interval of about 10 seconds between trials, during which the assistant read the instrument and set the indicating hand at zero. Immediately after the grip the subject called "right" or "left," as the case might be, and held out the instrument to the assistant to be read. The assistant did not call out the reading, but entered it in a form provided for the purpose. No particular effort was made to stimulate the subject, although before he began he was told to do his best. Usually someone else besides the assistant was present and watching him. Although no opportunity was given the subject to read the record on the dial carefully, he knew in a general way, from the position of the indicator, about what score he was making, and might be able to remember his performance from time to time, but no information was given him.

The particular instrument which was used was tested against standard weights to find if any corrections should be made in the reading.3 The results of these standard tests are shown in table 8. Two sets of readings were taken, that is, under temperature conditions of 25° C. and 13.5° C. The table is divided accordingly; there is a column for standard weights; one for the average reading from five entirely independent trials, and the average error, plus or minus, between the reading and the standard weights. The average error, particularly in that range which was commonly employed, which is from 35 to 70 kg., is usually less than ±0.2 kg. with the higher temperature, which was more nearly the condition used under actual experimentation, as the subjects desired to have the rooms in which they worked quite warm. It was seen that no such discrepancies or errors between standard weights and dynamometer readings were found in the case of this instrument as those recorded by Kohs, and therefore our readings are given without correction. With the exception of three occasions, the readings were always made by the same individual. These times were: the evening of December 19, Squad A; morning of February 3, after the walking experiment, Squad A; and the records which were taken in Springfield on May 21 and 22.

1 This instrument is manufactured by C. H. Stoelting Co., Chicago. The particular one employed belonged to the Massachusetts State Psychopathic Hospital, Boston, and was kindly loaned to us for the period of the experiment.

* See Whipple, Manual of Mental and Physical Tests, part 1, Baltimore, 1914, p. 100, for standard recommendation of procedure. The men were asked to remove any rings from their fingers during the teat.

2 Kohs, Journ. Exp. Psychol., 1917, 2, pp. 304 ff.

Table 8. - Calibration To Check The Accuracy Of The Hand Dynamometer Used In This Research