The preceding pages contain the temperatures of the districts through which the Expedition traveled, wherever I have been able to ascertain them, and also data for extending the lines of mean annual heat (isothermal), mean summer heat (isothæral), and mean winter heat (isocheimenal) across the continent. By comparing the sea-coast temperatures in Table II. and those of the shores of the great lakes in Table I. with those of places in France and Italy lying between the same parallels of 42°-45° north latitude, we perceive that the mean annual heat of Europe is from 8° to 15° Fah. greater than that of America at the same distance from the equator, while the summer heats differ only from 2° to 6°.* The inferior mean heat of America is therefore due principally to excessive winter colds, and this is decidedly the case in the interior. As the summer heats, however, regulate the culture of the cerealia and the growth of deciduous plants generally, the severe winters of America do not cause a scanty vegetation. From the 50th parallel northward the trees are frozen to their centres in winter; and, consequently, the development of buds and other vital processes which go on in the temperate climate of England, even in the coldest months, are completely arrested. This hybernation of plants increases in length with the severity and duration of winter which, generally speaking, augment in the interior of America with the latitude. The summer heats do not, however, decrease in the same ratio as we go to the north; on the contrary, the isothæral lines nearly follow the canoe route, and run to the northward and westward. The elevation of the prairie slopes has less influence in depressing the summer heat, than the nature of the soil and other causes have in raising it.

Experiments are still wanting whereby we may ascertain the ratio of the decrease of mean heat in America with the increase of altitude. In Table II. we find that, notwithstanding the eleva

*Dove's table in the Report of the Brit. Association for 1847 has furnished the means of making this comparison. The places compared were Alais, Arles, Bordeaux, Dax, Manosque, Marseilles, Montpellier, Pau, Puy, Tarascon, Toulon, Toulouse, Perpignan, Alba, Bologna, Cascina, and Lucca. Oleron of Bearn has the same altitude with Lake Superior, and Mount Louis is 4900 feet high. All of them lie within the parallels of 43°-45° except the last which is in 42° 50′ N. The maps of isothermal lines of this author express the general results of the study of the table referred to.

COMPARISON OF TEMPERATURES.

405

tion of Franklin Malone above Eastport of 645 feet, its mean temperature is greater; its interior position giving it an advantage in summer heat over the sea-coast, which its greater altitude does not destroy. If we refer to Dove's table, and contrast the temperatures of Mount Louis, which lies near the 43d parallel, with those of low country situations enumerated at the foot of the preceding page, we find a mean difference of temperature of one degree of Fahrenheit for 350 feet of altitude. A similar allowance for the elevation of the successive steps of the St. Lawrence basin would place in still greater prominence the rise of the isothermal lines, and more still that of the isothæral ones, as they recede from the Atlantic coast. There is, however, this difference in the climate of the summit of a mountain and of an elevated plateau, that in the former case we approach near the line of invariable temperature, and the summer heat therefore differs less from the mean of the year, and more from that of the plains, than on a plateau where the depression of mean temperature produced by elevation is due chiefly to winter colds, and in a small degree only to defect of summer heat.

From Table II. also we may learn that the mean temperature of the coast districts of the Pacific is greater than that of the Atlantic countries, and, at the same time, more equable; the difference between the hot and cold months being less. We find in it an expression of the general fact, that the west coasts of continents are warmer than the east ones; and as Montreal and Fort Vancouver lie nearly in the same latitudes and at the same altitudes above the sea, and both are far enough removed from the coast to be beyond the direct influence of the sea breezes, columns 4 and 6 furnish the means of eliciting many of the peculiarities of climate on the two sides of the continent. Instead of four or five months of continuous snow and ice which Canada may be said to enjoy, for it is the season of general enjoyment, Oregon has an open, rainy winter, with little frost or snow; but, at the same time, à summer of less power.

Table V. exhibits even greater differences in the Pacific and Atlantic climates in a higher parallel. The course of the ocean currents, and the interposition of the peninsula of Alaska and its prolongation by the Aleutian chain of islands, protect the west coast of America from the masses of drift ice which in the same latitudes encumber and chill the Labrador coast for the most of the year. Even in the polar regions the west coasts have milder climates. Table X. shows, as far as it goes, that the mean temperature of the west coast of Greenland exceeds that of places on the continent, up to the 150th meridian, though the summer on the coast is greatly colder than that of the interior. By the study of Table XII. we learn that in the polar seas the summer heats vary little, as we might

expect from the constant presence of ice; but the annual mean seems to decrease generally with the latitude, the only exception being that of Wolstenholme Sound, in which we have a confirmation of the greater mildness of the west Greenland coast. In the high latitudes the mean heat of the three winter months does not differ greatly in different years; but in some seasons one of these months, in others another, is the coldest, the temperature being ruled greatly by the prevailing winds.

Generally speaking, the mean annual temperature of places in the interior of North America falls within a degree or two of the mean heat of the two months of April and October. The mean temperature of the whole surface of the earth is, according to Dove's calculations, 58.2° Fah., being 54° in January and 62° in July. For such a mean annual heat in America we must descend to the 34th parallel of latitude; but the July heat of 62° Fah. extends northward to the Mackenzie.

The intense winter colds in the high latitudes are apparently in a great measure owing to the active nocturnal radiation into the clear blue sky. The observatory, which was a small log building without a fire-place, furnished us with the means of judging how much greater the depression of temperature in the night was in places exposed to the sky, than in those covered in.

The daily curve of atmospheric temperature for the three winter months was a bold and nearly regular hyperbolic curve, of which the mean was -25.2° Fah., the maximum -22.2° Fah., and the minimum-26.7° Fah. The maximum occurs at 1h. 18m. P.M., and the mean line is crossed by the curve at 9h. 19m. A.M., and 6h. 29m. P.M.; the lowest temperature being reached at 7 A.M. The ascending branch of the curve, therefore, corresponds to an interval of 6h. 18m., and the descending one to 17h. 42m. During 14h. 51m. the temperature is below the mean, and it is only 9h. 9m. above it, which indicates a tolerably bold curve in the day, and a nearly horizontal course in the night.

In the observatory the mean for the same period was only -15.91° Fah., and the range no more than 0.97°. The maximum occurred at 6 P.M., being retarded 43 hours; and the minimum at 10 A.M. being delayed 3 hours. For most of the night the temperature was above the mean-such being the effect of the interposition of the building between the thermometer and the blue atmosphere. The walls of the observatory, it is necessary to remark, were by no means air-tight, and the door was opened at least once an hour in the day, and sometimes, especially on term days, much oftener. There was, consequently, a considerable and frequent admission of

*The correction for the error of the thermometer at low temperatures used in Tab. X. col. 9, was not applied to these numbers.

EFFECTS OF SUNLIGHT ON VEGETATION.

407

the external air; and, on the other hand, during the experiments on magnetic intensity, the heat of the observer's body had an evident effect in raising the temperature of the room.

I had intended to have instituted a series of observations, with Sir John Herschell's actinometer, on the nocturnal radiation, and also on the momentary intensity of the direct rays of the sun; but the instrument was unfortunately broken on the voyage. The Edinburgh New Philosophical Journal for 1841 contains the results of observations made at Fort Franklin, with the black bulb thermometer, on the heating power of the sun's rays, and I renewed these observations at Fort Confidence; but, as they were not carried on later than April, they furnish no information respecting the power of the sun in the months in which the processes of vegetation are active. As the black bulb thermometer indicates the accumulative effect of the sun's rays, it seems to be a useful instrument for ascertaining the heating power of the sun on the stems and larger branches of trees, at least, if not also on their leaves and on herbaceous plants. The hybernation of trees ceases long before the temperature of the atmosphere is sufficient to restore activity to the vegetative processes, and before the earth, still enveloped in its snowy covering, has felt the influence of returning spring. is evidently mainly or wholly due to the sun's light direct or reflected; and perhaps its rays as reflected from the pure and glassy surface of the snow, after the days have increased considerably in length, may have the same powerful effect on the forest that, according to Professor Forbes, they have on the black-bulb thermometer. For some time after the trees have begun to thaw by day, they freeze again in the night; and in more southern localities, where the sugar-maple grows, the sugar makers are well acquainted with the fact that a hard frost arrests the flow of the sap in the night. Should a hot day, however, follow such an occurrence, the flow is more abundant than ever, the short rest seemingly increasing the irritability of the organs by which the sap is eliminated and circulated.*

This

* As I was revising this sheet, Sir William Hooker favored me with an extract from the journal of Mr. Berthold Seeman, botanist of the Herald, part of which follows: "During our stay at Port Clarence, in September, 1850, I made several experiments to ascertain the depth to which the thaw penetrated the soil: the result varied; in some places it did not descend above two feet into the earth, while in sandy places the ground was free from frost to the depth of four or five feet. The season was much colder than in 1849, the sea more loaded with ice, and the terrestrial vegetation less vigorous."

No. III.

ON THE GEOGRAPHICAL DISTRIBUTION OF PLANTS IN THE COUNTRY NORTH OF THE 49TH PARALLEL OF LATITUDE.

Generic and Specific Forms of Plants decrease in Number as the Latitude in creases. Analogy between Altitude and Increase of Latitude.-Culture of the Vine. Of the Cerealia.-Maize.-Wheat.-Oats.-Barley.-Potatoes.-Botanical Districts.-Their Physiognomy.-Woodland District.-Barren Grounds.Prairies.-Rocky Mountains.-Sitka.-Polar Plants.-Arctic Zone.-Trees and Shrubs.-Table of Distribution of Species in three several Zones.-Carices.

THOUGH the isotharal lines, when the term is restricted to the mean temperatures of the three summer months of June, July, and August, run from Lake Superior northward to the Mackenzie, yet the short duration of the summer on the banks of that river, and the occasional frosts in June and August, and in some years even in July, render the climate unsuitable for numbers of vegetables which flourish in the northern districts of the United States. Many trees, shrubs, and perennial roots can be frozen without injury if the frost be continuous throughout the winter; and they acquire so much irritability in their hybernation, that the stimulus of perpetual though less fervid day within the arctic circle causes them to perform the functions of foliation and fructification with a rapidity unknown in more temperate regions. Other plants which need longer time to perfect their fruit or woody fibre, terminate in succession according to their several constitutions as the latitude increases. Their place is only partially supplied by other species, which have in like manner their equatorial and polar limits. These are not, however, so numerous as those which die out, there being no rule more general than the decrease of generic and specific forms in passing from temperate zones to arctic or polar ones.

There is a similarity in many respects between the vegetation of alpine tracts and that of high latitudes, but not an identity, the condition of the two regions differing in some essential particulars. No more apt illustration of this fact is needed than that adduced by Meyen, of Titicaca. This alpine lake, situated on the plateau of Chuquito in southern Peru, at the elevation of 12,700 feet, is surrounded by a rich and beautiful vegetation, which flourishes under a perpetual spring. On its banks a populous community, inhabiting magnificent cities, is supported by a fertile soil, yet trees are want