BILATERAL ASYMMETRY OF FUNCTION. 105 exact opposite occurred, the excess being on the side of the non- preferred hand. This observation can only be made within limited angular distances. If the pins are too far apart, so that, when the eyes are fixed on one, the other is seen far in the field of indirect vision, the error is more irregular, and, so far as our limited figures cover this point, greater in the same sense as above. These observations must thus be restricted to small angu- lar distances. B. In a second series of observations, two rods or rulers, similar to that described above, were fastened, one in a perpendicular, the other in a horizontal position, crossing each other in the middle at the height of the shoulders. These four quadrants we num- bered from the side of the recorder, the upper quadrant at his left being Xo. 1, and around with the hands of a watch to quad- rant four at his left hand below. The person under observation assumes the primary position, with erect body and eyes closed, extending both arms horizontally and nearly straight, and laying his two index fingers together in angle of the first quadrant. The fingers being then moved simultaneously as in A. I., the left passes up along the perpendicular rod, while the right finger moves through a distance intended to be equal, outward to the right along the horizontal rod ; both distances being noted from the other side as before, and the observer passing later to the other quadrants in order. Here the effect of the gravity of the arms was almost always predominant, and even in the longest series of observations, after the experimenter had learned his error, was never compensated for. Whichever arm moved up made a less, and whichever arm moved down made a greater, excursion than the arm which moved horizontally at the same time. Yet if the right arm of a right-handed person moved up and the left out horizontally, as in the second quadrant, the excess of the latter over the upward movement was less than if the left went up and the right outward, as in the first quadrant, while in the two lower quadrants the excess of the downward over the hori- zontal movements was greater when made by the right than when made by the left hand. If, then, as seems reasonable to infer, our knowledge of the plane surface in which this cross lay were dependent upon the muscular and innervation-sensations in- volved in these movements alone, an inch in this surface would seem much longer above the level of the shoulders than below them, and somewhat larger to the left of the median line than to the right, for a right-handed person: a supposition, of course, only abstract, because, in fact, we leam to judge of all the space within reach of our hands and arms by the movements of both arms in both directions, and especially by the eyes. When the hands were alternately weighted, the relative excess of the down- ward, and the deficiency of the upward movements was easily increased. Many series of successive movements, corresponding to A. II.,
