the curvature in the record is slight, and practically negli- gible in the middle portion of 4 cm. The dimensions given allow a magnification of ten thousand times. A far more compact apparatus is made with 15 cm. length of levers. This gives a magnification of a thousand times.
AUTOMATIC RECORD OF THE RATE OF GROWTH.
Another great difficulty in obtaining an accurate record of the curve of growth arises from the friction of contact of the bent tip of the writing lever against the recording surface. This I was able to overcome by an oscillating device by which the contact, instead of being continuous, was made intermittent. The smoked glass plate, G, is made to oscillate, to and fro, at regular intervals of time, say one second. The bent tip of the recording lever comes periodically in contact with the glass plate during its ex- treme forward oscillation. The record would thus con- sist of a series of dots, the distance between successive dots representing magnified growth during a second.
The drawback in connection with the obtaining of record on the oscillating plate lies in the fact that if the plate approaches the recording point with anything like suddenness, then the stroke on the flexible lever causes an after-oscillation ; the multiple dots, thus produced, spoil the record. In order to overcome this, a special contrivance is necessary, by which the speed of approach of the plate should be gradually reduced to zero at contact with the recording point. The rate of recession should, on the other hand, continuously increase from zero to maximum. The recording point will in this manner be gently pressed against the glass plate, marking the dot, and then gradu- ally set free. It was only after strict observance of these
