In the next trial much younger cotyledons were similarly treated, but were exposed to a rather obscure lateral light.  After some hours, a bowed cotyledon, .3 inch in height, had an open furrow on the shaded side .04 inch in breadth; another cotyledon, only .13 inch in height, had left a furrow .02 inch in breadth.  But the most curious case was that of a cotyledon which had just protruded above the ground and was only .03 inch in height, and this was found to be bowed in the direction of the light to a depth of .2 of an inch beneath the surface.  From what we know of the impermeability of this sand to light, the upper illuminated part in these several cases must have determined the curvature of the lower buried portions.  But an apparent cause of doubt may be suggested:  as the cotyledons are continually circumnutating, they tend to form a minute [page 476] crack or furrow all round their bases, which would admit a little light on all sides; but this would not happen when they were illuminated laterally, for we know that they quickly bend towards a lateral light, and they then press so firmly against the sand on the illuminated side as to furrow it, and this would effectually exclude light on this side.  Any light admitted on the opposite and shaded side, where an open furrow is formed, would tend to counteract the curvature towards the lamp or other source of the light.  It may be added, that the use of fine moist sand, which yields easily to pressure, was indispensable in the above experiments; for seedlings raised in common soil, not kept especially damp, and exposed for 9 h. 30 m. to a strong lateral light, did not form an open furrow at their bases on the shaded side, and were not bowed beneath the surface.  Perhaps the most striking proof of the action of the upper on the lower part of the cotyledons of Phalaris, when laterally illuminated, was afforded by the blackened glass-tubes (before alluded to) with very narrow stripes of the varnish scraped off on one side, through which a little light was admitted.  The breadth of these stripes or slits varied between .01 and .02 inch (.25 and .51 mm.).  Cotyledons with their upper halves enclosed in such tubes were placed before a south-west window, in such a position, that the scraped stripes did not directly face the window, but obliquely to one side.  The seedlings were left exposed for 8 h., before the close of which time the many free seedlings in the same pots had become greatly bowed towards the window.  Under these circumstances, the whole lower halves of the cotyledons, which had their summits enclosed in the tubes, were fully exposed to the light of the sky, whilst their upper halves received exclusively or chiefly diffused light from the room, and this only through a very narrow slit on one side.  Now, if the curvature of the lower part had been determined by the illumination of this part, all the cotyledons assuredly would have become curved towards the window; but this was far from being the case.  Tubes of