For mean errors:

(1) the curve for head-forward and head-backward closely corresponds to l.h.e.c. record; the errors increase by 50% with increase of speed-rate, suggesting that

(a) visual control is negligible, as far as seeing the moving pencil is concerned;

(b) control of head for forward equals that for backward movements.

For constant errors:

(1) there is underruling at high speeds because of the usual physiological limitations, and this is more marked for head backward results, suggesting that

(a) spatial relations are obtained, when the apparatus is visible, that tend to correct underruling, or

(b) an extra inhibitory effect, due to lack of knowledge of spatial relations, is added to the normal physical shortening and the subject moves the head a less distance than is naturally possible; or

(c) the muscular control is less complete for movements of the head backward.

For individual cases we find: