of the Polar Regions. 67 



cene epoch is too nearly approximate to our own to allow of 

 our attributing to it, with any probability, the diflference of 

 temperature now indicated. 



It seems to us to follow, from the preceding considerations, 

 that it is from the study of phenomena of another kind that 

 we must obtain the solution of the problem that we seek. Let us 

 first examine, in the series of cosmical phenomena, the question 

 of the changes which may have taken place in the position of the 

 earth relatively to the sun. From the point of view which now 

 engages our attention, a great importance has been quite re- 

 cently attached to the periodical modifications of the eccen- 

 tricity of the ellipse which is annually traversed by the earth. 

 The form of this ellipse is modified within certain limits in 

 the com-se of ages. At present it approaches a circle j in 23,900 

 years its eccentricity will have attained its minimum ; then 

 the orbit will again tend to acquire a more elongated form. 

 The mean distance from the earth to the sun is 91,400,000 

 miles ; when the eccentricity of the ellipse is at its maximum 

 it has -yV of this length, when it is at its minimum -^\-^ ; in 

 the former case the earth departs from the sun 14,500,000 

 miles more than in the second case. At present the linear 

 value of the eccentricity is three millions of miles. It must 

 also be observed that, at present, the earth is nearest to the 

 sun (at the perihelion) during the winter of the northern 

 hemisphere, whilst in the summer it is fm-thest from it (at the 

 aphelion). Noav the relative position of the line of apsides 

 and of that of the solstices is likewise subjected to a movement 

 of revolution which is accomplished in 21,000 years. In 

 about 10,000 years the summer of the northern hemisphere 

 will fall at the period when the earth is nearest to the sun, 

 and its winter at that when it is most distant. Of course the 

 contrary will take place in the southern hemisphere. 



It is assumed that dm-ing the periods when the eccentricity 

 of the orbit of the earth approaches its maximum, when the 

 perihelion coincides with the winter solstice, the northern 

 hemisphere must enjoy a shorter and warmer winter, whilst 

 the summer is longer and cooler. Dm'ing this time the oppo- 

 site is the case in the southern hemisphere. Its winter is 

 longer and colder, its summer hotter and shorter, because the 

 winter of this hemisphere corresponds with the greatest dis- 

 tance from the sun. Mr. Croll supposes that during this long 

 and cold winter so great a quantity of ice must have been 

 accumulated, that the summer following it (which, although 

 hot, was short) had not the power to melt it entirely, and that 

 it is at this epoch that we must place iho, glacial period. In 

 the northern hemisphere, on the contrary, there would have 



5* 



