GEOLOGICAL SEASONS. 187 



with the earth's aphelion once or more during the con- 

 tinuance of a period of high eccentricity. At the present 

 time the winter solstice of the northern hemisphere oc- 

 curs in perihelion. When brought to occur in aphelion 

 during a cycle of extreme eccentricity, the earth would 

 be 8,641,870 miles farther from the sun in winter than 

 at present. This difference would cause the sun's inten- 

 sity to be one-fifth less during winter than at present. 

 It is true that it would also be one-fifth greater during 

 summer; and thus the annual constant of the solar heat 

 would not be diminished. To speak more precisely, it 

 would actually be increased by one three-hundredth part, 

 since the annual amount of heat is inversely proportional 

 to the minor axis of the earth's orbit. It is also true 

 that while the sun's intensity during the northern winter 

 would be diminished one-fifth, the duration of the season 

 would be prolonged forty-four days beyond its present 

 length, and would be thirty-six days greater than the 

 duration of the summer. Thus not only would the win- 

 ter heat be diminished, but the diminution would be pro- 

 longed. This would, indeed, secure the same absolute 

 aggregate of winter heat as at present; and this conclu- 

 sion is as far as Mr. Croll's predecessors went in the in- 

 vestigation of this problem. The total amount of winter 

 heat being the same, its total effect, they argued, would 

 be the same. Mr. Croll's merit consists in taking into 

 account the effect of a diminished daily intensity, and of 

 the extension of this through a longer period. In all 

 climatic investigations, as is shown in another chapter of 

 this work, the means of short periods are quite as im- 

 portant as the means for long periods. It is the extreme 

 cold of winter which conditions the growth of vegetation, 



