420 BOTANICAL GAZETTE [JUNE 
be, however, that there are times when the temperature barely 
falls to the death point, and then only for a very short time just 
before sunrise, when the bud-scales may save the life of the bud 
by preventing a temperature fall of from 0.5° to 1° for a very short 
time, and this little we must admit they are capable of doing. But 
we must consider that the period when the young shoot is covered 
by the extending scales, in America at least, lasts only about three 
or four days at the most, while the probability of a fall just to the 
critical temperature during this short period is indeed very slight. 
There may be no frost at all or there may be a very severe one. 
Only in the case of such a slight frost just reaching the critical tem- 
perature could the scales be of benefit, and even this injury would 
not be fatal to the tree, since another crop of accessory buds can 
grow out in a short time. The chance to function is therefore very 
slight, and the effect could not compare in importance to the plant 
with the benefits obtained according to the above theory, because 
in the absence of scales nearly all buds of whatever crop would 
run great danger of being killed by loss of water. It is not reason- 
able to suppose, therefore, that the benefit derived by modifying 
temperature is sufficient to have played any great part in caus- 
ing natural selection to evolve such an elaborate structure for this 
purpose. 
Relation oj bud-scales to mechanical injury of the bud. 
The idea that the bud-scales serve to protect the delicate young 
tissues within from mechanical injury is of course not new.*' In 
fact, probably everyone feels that this must be, if not the most impor- 
tant function, at least a prominent factor in the work of protection. 
Nevertheless the subject seems to have received little attention in 
physiological works. 
The young shoots of our definite-growing trees as they exist 
through the winter are very tender organs, composed mainly of 
thin-walled parenchyma. In most cases the epidermis is still thin, 
no fibrous or other supporting tissue has been developed, and the 
vascular bundles contain only spiral vessels, the result being that 
they are very brittle and capable of resisting only the slightest con- 
cussions. The effect of these abrasions is besides very much inten- 
3« Groom, P., Bud protection in dicotyledons. Trans. Linn. Soc. II. 3:255- 1893- 
