skip to content

Darwin Correspondence Project

From James Croll   [2 December 1868]1

The glacial condition of climate seems to result from a complication of physical causes brought into operation by an increase in the excentricity of the earth’s orbit. When the excentricity is about its superior limit, the combined effect of all those physical causes is to lower to a very considerable extent the temperature of the hemisphere whose winters occur in aphelion, and to raise to nearly as great an extent the temperature of the opposite hemisphere, whose winters of course occur in perihelion. Owing to the precession of the equinoxes and the motion of the perihelion the winter solstice moves from perihelion to aphelion and from aphelion to perihelion making a revolution in a period which is very irregular varying from 20,000 to 30,000 years* The winters will therefore pass from aphilion to perihilion or from perihilion to aphilion in about 10,000 or 15,000 years. Consequently during the glacial epoch the ice would be transferred from the one hemisphere to the other every 10,000 or 15,000 years.

Take the case of the N. hemisphere; when its winter solstice was half way between perihilion and aphelion and moving in the direction of the latter, the cold would begin to increase and the ice accumulate on the N. hemisphere as the solstice point approached nearer and nearer to the aphelion and even perhaps for a considerable time after it had passed the aphilion; but as it began to draw round to the middle point in its way back from aphelion to perihelion the cold would decrease and the ice melt. This gives a period of ice of about 10,000 or 15000 years.

As the solstice point approached the perihelion the climate would continue to grow warmer and warmer till the perihelion was reached. The winters at this time would not only become much warmer than they are at present but would be also more equable; for owing to the nearness of the sun during winter the winters might perhaps be as warm as the summers.

You will see from the tables given in pages 104–106 that when the excentricity reaches a very high value owing to the extreme slowness with which it changes it continues for a long course of ages in that condition.2 Suppose it to continue for say 100000 years or so; the ice would be transferred from the one hemisphere to the other several times during that long period   The cold of the glacial epoch must have therefore been broken up by a number of warm periods of 10000 or 15000 years duration each. The same thing hold true of Southern hemisphere. I have considered the glacial epoch to be that long period of 160000 years included in Table IV. p. 106. This epoch must have consisted according to theory of a series of cold and warm periods extending to both hemispheres, but of course the warm periods of the one hemisphere correspond to the cold periods of the other and vice versa

The rate of precession is subject to so many disturbing causes that when we go back to a period of one or two hundred thousand years it is impossible to determine whether at any particular time our winter was in perihelion or aphelion. This however is of no importance to the question

My own opinion is that the trees which are found in Melville Island and other arctic spots grew during some of the warm periods of the glacial epoch when in all probability Greenland would be enjoying a tolerably mild climate.3 The reason why we have now on our existing continents so little remaining of what flourished during those periods is easily perceived.

The explanation how it comes about that when the excentricity is at a high value the climate is cold when the winters are in the Aphelion, and warm when in the perihelion involves a good many purely physical point   I have only been able as yet to refer briefly to a few of them in former papers.

1. When the winter is in the aphelion and the excentricity at a high value the winter half year is much longer than the summer. For example about 850000 years ago if our winter was then in aphelion it would actually be 3412 day longer than the summer (see Table III Column V. p. 105)4 The snow accumulating period would therefore be more than a month longer than the snow melting period.

2 The winter would not merely be longer but also much colder owing to the suns greater distance. The winter would be then more than 40oF. colder than at present (Column VII)   We would then have no rain during winter—nothing but snow.

3. Although the sun would be much nearer during summer than at prest the summers would not on that account be warmer than they are now—but much colder. The physical cause of this is stated in pp 54. 67. 108.5

4 But the most powerful of all the causes is that resulting from the influence that excentricity has upon ocean currents. The Dynamical Theory of Heat enables us now to estimate the amount of heat effects in absolute measure. By applying this meathod to the determination of the amount of heat conveyed by currents, we arrive at results that are truly remarkable. For example we find that the quantity of heat conveyed from the tropical regions by the Gulf Stream alone is actually nearly equal to all that is received from the sun by the entire Arctic regions even supposing that none of the sun’s rays were cut off in passing through the atmosphere.6

It has been shown that the effect of excentricity would be to turn the entire current into the Southern Ocean during a glacial period on the Northern hemisphere. So far as I am aware the proof adduced has not been disputed (see on this point pp 16. 64)  

The effect that this would have in lowering the temperature of the Northern hemisphere and raising the temperature of the Southern would be very great. see pp 69. 70. 153. 154.7 But it would not be the course of Gulf Stream alone that would be altered for as the medial line between the trades (p 64) would not be at the equator as at present, but would lie at a very considerable distance to the south of the equator the warm surface water at the equator would be pushed upon the southern hemisphere. The heat would all go to raise the temperature of the southern hemisphere. The water after parting with its heat would sink and flow back upon the northern hemisphere as a cold under current. The cold water would flow to the northern hemisphere across the equator as an under current— the surface water would cross the equator to the southern hemisphere as a warm current. I have some unpublished calculations on hand regarding the effect that this would have in lowering the temperature of the Northern hemisphere and raising the temperature of the Southern. The results are startling.

Sir Charles Lyell in his Principles comes to the conclusion that the hemisphere whose winters were in perihelion would not have its temperature much raised although the other hemisphere would be very much lowered.8 I am sorry I am obliged to differ from him on this point. I may mention however that at the time Sir Charles wrote no calculations had been published regarding the positive amount of heat conveyed by the Gulf Stream.9

(5.) If the effect that the great preponderance of the N.E. trade winds over the S.E. (resulting from the greater difference between the temperature of the N. Pole and the equator, than between the S pole and equator) would be to impel the warm water at the equator to the south, the effect of the strong upper current—return trade—would be to carry the aqueous vapour formed at the equator to the north. The upper current on reaching the cold ice-sheet would deposit its moisture in the form of snow. This would be particularly the case during summer when the earth would be in perihelion and the heat at the equator excessive.

It is a curious circumstance that all the agencies in operation should tend to one effect viz the production of snow on the one hemisphere and the melting of it on the other. This seems to be the grand secret of the Glacial Epoch. The heat was all pushed to the one side of the equator and the cold and moisture to the other side. Even the very nearness of the sun during summer tended to increase the amount of snow falling.

6 If the ice accumulate on the one hemisphere and melt on the other this will shift the earth’s centre of gravity see pp 137–142 also p 189. If the earths centre of gravity is shifted the sea must shift along with it. Consequently a submergence of the land during the Glacial Epoch is a necessary result

7. In pp 194, 195 you will find some speculations as to a cause which might have made the ice move over the face of N America. You must however take them for what they are worth.

*see page 163 footnote10

CD annotations

1.1 The … orbit. 1.2] scored pencil
3.2 The winters … summers. 3.5] scored pencil
4.10 but … vice versa 4.11] scored pencil
11.1 But … currents. 11.2] scored pencil
12.2 So … great. 13.2] scored pencil
14.3 I am … Stream. 14.6] scored pencil
Verso of last page: ‘James Croll—Geolog. Survey of Scotland, Edinburgh | Abstract of views on Glacial Periods—’ ink


The date is established by the relationship between this letter and the letter from James Croll, 2 December 1868.
Croll’s page numbers evidently refer to his own pagination of his collected articles, which he had put together in book form and sent to CD along with this abstract of his views on climate change (see letter from James Croll, 2 December 1868 and n. 2). Croll probably refers to tables contained in Croll 1868, pp. 148–50.
Melville Island, the largest of the Parry Islands, is in the West Baffin region of the Canadian Arctic. Miocene fossil plants had been collected on Melville Island (Lyell 1867–8, 1: 225; see also Croll 1867b, p. 429).
See Croll 1868, p. 149.
See Croll 1868, p. 152. The other references have not been identified, but see Croll 1864, pp. 123, 131.
For calculations of the amount of heat conveyed by the Gulf Stream, see Croll 1867b, pp. 433–4.
Croll’s references have not been identified, but see Croll 1864, pp. 135–7, and Croll 1867c, pp. 122–3.
See Lyell 1867–8, 1: 268–73.
See n. 6, above.
Croll’s references have not been identified, but see Croll 1865 and 1866.


Glacial climates.

Letter details

Letter no.
James Croll
Charles Robert Darwin
Source of text
DAR 50: E3–8

Please cite as

Darwin Correspondence Project, “Letter no. 6489F,” accessed on 26 April 2019,

Also published in The Correspondence of Charles Darwin, vol. 16