Notes on the comparative hardiness of plants are always valuable. When any one thing has been found to endure every variety of unlucky experiences, it is so much in its favor. But it must not be forgotten that the discoveries of late years put a new interpretation on these facts Plants, woody plants particularly, die during winter from the evaporation of their juices, and not because their cells have burst by liquid expansion; hence we have to take evaporating influences into consideration quite as much as the state of the thermometer. Indeed, the thermometrical degrees are of no account, only so far as they influence evaporation. All other things being equal, there would be more evaporation at 20° than at 30°, and more at 10° than at 20°. But dry wind has an influence, and, as clearly shown by Deherain, Anders and others, light has an influence. Two trees of the same species may not be fifty yards apart; one may be in the eddy of a wind current, and the other protected by a single bush from both wind and bright light - the former get killed and the latter escape. The next year it might be reversed. The dry, cold current coming from another quarter, the favored one last year would be the one to suffer next.

Besides this, the ability of the roots to supply an excessive drain on the liquid reservoirs has much to do with hardiness, and the plant which has been lucky enough to send down a good tap root, has a great advantage in the war with the elements. But even the past has an influence. At the time the leaves begin to fade in the fall - that is to say in the early autumn season - there is less moisture in wood structure than at any other time; but from the time the leaves begin to fade the wood begins to lay up a store of moisture against the drafts of winter. If the cells have been injured by excessive drouths, by fungus attacks, or by any other weakening influences, they cannot store up as much as they would otherwise, and such a plant will die before another which has had a better chance, though every other circumstance is equal.

We thus see that though the condition of the thermometer has much to do with hardiness, it is only so far as it influences evaporation; and there are numberless other conditions which influence evaporation besides.