E. P. Culverwell — Theory of the Ice Age. 7 



received from the Gulf Stream bears to the winter sun-heat, 

 and then he could have estimated the percentage reduction in 

 the ivhole heat supply due to the known reduction in sun heat. 

 In addition to this, it was quite out of date for him to take a 

 proportionate reduction at all. Even when the first edition of 

 " Climate and Time " was published, physicists knew that the re- 

 duction was far less, and four years before the second edition was 

 published, the law of the 4th power, used later on in this article to 

 show how Dr. Croll might have worked the problem, had been 

 published by Stefan. But Dr. Croll was content to found his theory 

 on an assumption of Herschel in his " Outlines of Astronomy," 

 § 369 a, although his book was written to show that Herschel was 

 quite wi'ong in his conclusions on this very matter. 



To show how far even the most extreme midsummer and mid- 

 winter temperatures are from being proportional to the sun-heats 

 received at those times, take Yakutsh in Siberia, supposed to be the 

 most extreme of terrestrial climates. The excess of its summer 

 temperature over the — 239° F. is about 309° F., and of its winter 

 temperature about 199° F., but the midsummer sun-heat is to 

 the midwinter sun-heat as 5800 is to 199, or as 309 is to 11. 

 Hence we may draw two conclusions : if the excess of mid- 

 winter temperature over — 239° F. be due to the midwinter 

 sun-heat, then, on Croll's theory, the midsummer temperature ought 

 to be 5800° F. above the — 239° F. ; or if, on the other hand, the 

 summer excess of 309° F. be due to the summer sun-heat, then the 

 midwinter temperature ought, if only dependent on sun-heat, to be 

 only 11° F. over the — 239° F., i.e. it ought to be — 228° F., and 

 the rest of the 199° F. excess over the — 239° F. must be due to 

 heat derived from other regions of the earth. If we apply this 

 reasoning to Great Britain the difference will be more striking still, 

 because there the midwinter excess over the — 239° F. differs little 

 from the midsummer excess. If, instead of taking Croll's sun-heat 

 on the single days of least and most insolation, we take Ball's average 

 daily summer and winter sun-heat, the result is not, perhaps, so 

 grotesquely different from the assumption, but it is sufficiently 

 different to show the absurdity of the supposition ; for the ratio 

 of the average daily sun-heat in summer to that in winter is about 

 1000 to 199, or 309 to 62. Hence, if the winter temperature be due 

 to the sun-heat, the summer temperature should be 1000° F. over 

 — 239°, or, say, 760° above the zero of the Fahrenheit scale ; or, if 

 the summer tempei'ature be that due to sun-heat, then the winter 

 temperature ought to be — 239°-l-62° F., or — 177° F., instead of 

 about — 40° F. 



The plan which I have adopted for measuring the fall in tempera- 

 ture is as follows : — It is not difficult to ascertain what parallel of 

 latitude noio receives the same lointer sun-heat as parallel 50° received 

 in the period of great eccentricity ivhen the winter loas in aphelion. 

 We can also do this for the parallels of 40°, 60°, 70°, 80°, 

 and 90°. I have made the necessary calculations on three different 

 suppositions. First, following Croll's idea, I find what parallels 



