whilst sinking downwards in the evening twist round, so that the upper surface is turned inwards, and is thus better protected than the outwardly turned lower surface.  Nevertheless, it was always the upper surface which was more blackened than the lower, whenever any difference could be perceived between them; but whether this was due to the cells near the upper surface being more tender, or merely to their containing more chlorophyll, we do not know.

Melilotus officinalis.—­A large pot with many plants, which had been kept during the winter in the greenhouse, was exposed during 5 h. at night to a slight frost and clear sky.  Four leaves had been pinned out, and these died after a few days; but so did many of the free leaves.  Therefore nothing certain could be inferred from this trial, though it indicated that the horizontally extended leaves suffered most.  Another large pot with many plants was next exposed for 1 h., the temperature on the surrounding grass being lower, viz., -3o to -4o C. Ten leaves had been pinned out, and the result was striking, for on the following morning all these were found much injured or [page 291] killed, and none of the many free leaves on the several plants were at all injured, with the doubtful exception of two or three very young ones.

Melilotus Italica.—­Six leaves were pinned out horizontally, three with their upper and three with their lower surfaces turned to the zenith.  The plants were exposed for 5 h. to a clear sky, the temperature on ground being about -1o C. Next morning the six pinned-open leaves seemed more injured even than the younger and more tender free ones on the same branches.  The exposure, however, had been too long, for after an interval of some days many of the free leaves seemed in almost as bad a condition as the pinned-out ones.  It was not possible to decide whether the leaves with their upper or those with their lower surfaces turned to the zenith had suffered most.

Melilotus suaveolens.—­Some plants with 8 leaves pinned out were exposed to a clear sky during 2 h., the temperature on the surrounding grass being -2o C. Next morning 6 out of these 8 leaves were in a flaccid condition.  There were about 150 free leaves on the plant, and none of these were injured, except 2 or 3 very young ones.  But after two days, the plants having been brought back into the greenhouse, the 6 pinned-out leaves all recovered.

Melilotus Taurica.—­Several plants were exposed for 5 h. during two nights to a clear sky and slight frost, accompanied by some wind; and 5 leaves which had been pinned out suffered more than those both above and below on the same branches which had gone to sleep.  Another pot, which had likewise been kept in the greenhouse, was exposed for 35 — 40 m. to a clear sky, the temperature of the surrounding grass being between -3o and -4o C. Nine leaves had been pinned out, and all of these were killed.  On the same plants there were 210 free leaves, which had been allowed to go to sleep, and of these about 80 were killed, i.e. only 38 per cent.