26 
logarithms of radius vector, change of midwinter heat com- 
pared with a mean distance, and results for Great Britain. 
88 
*0415 
302 ° 47 ' 
•01564 
16°-18 
48°-41 
87 
•0544 
135 ° 7 ' 
-•01572 
- 17°"07 
15°-16 
86 
*0631 
326 ° O ' 
•01679 
18°-23 
50°-46 
85 
•0664 
156 ° 22 ' 
- -00980 
- 10°-64 
21°-59 
84 
•0622 
346 ° 46 ' 
•00782 
8°-49 
40°-72 
83 
•0558 
177 ° 42 ' 
- -00039 
- 0°-42 
31°-81 
82 
•0432 
10 ° 16 ' 
- -01058 
— 11°"49 
20°-74 
24 
•0345 
71 ° 2 ' 
- -01393 
- 15°-12 
17°-11 
23 
•0433 
264 ° 21 ' 
•01914 
20 o> 78 
53°-01 
22 
•0450 
101 ° 47 ' 
- -01868 
- 20°-28 
ll °-95 
21 
•0517 
301 ° 35 ' 
•01988 
21°-58 
53°-81 
20 
•0512 
138 °- 22 ' 
- -01389 
- 15°-08 
17°-15 
19 
•0480 
336 °- 46 ' 
•00815 
8°-85 
41°-08 
18 
•0432 
177 0 > 38 ' 
+ -00004 
°-04 
32°-27 
17 
•0399 
19 ° - 3 ' 
- -00499 
- 5°-42 
26°-81 
10 -0306 
In A.D. 1800 the eccentricity is •01G78, the anomaly 98°, 
log. of radius vector in midwinter *00623, the increase + 6 0, 77 
and 39°— G°* 77 == 32°*23 is the midwinter heat of Great 
Britain, in a circular orbit, to be added to tlio degrees in 
col. 5, to obtain the midwinter heat on Mr. Croll’s hypothesis, 
after duo corrections. 
4G. Thus it appears, when the principle of Mr. Crolhs calcu- 
lation is admitted, and necessary corrections are introduced, 
the midwinter depression, or increase of cold in Great Britain, 
at his earlier date, B.C. 850,000, would not bo 45°*3, but 
only 10 o, 6; and that in B.C. 210,000 there would not be a 
decline of 37°*7, but a rise of 21°*6. At B.C. 220,000 there 
would be a decline of 20 o, 3 ; and this is fourteen degrees 
less than the amount in his theory. And when we obsorvo, 
further, that the same principle would involve the consequence, 
that southern winters should now be 13° colder than at tho 
same latitudes in the northern hemisphere, while there is 
actually only a very slight difference, the disproof of the 
hypothesis seems tolerably complete. 
47. The way to x’estore some semblance of truth to tho 
theory is to apply it, not to the periods in round numbers in 
the table, but to intermediate dates, when the solstice was 
really in the aphelion. This is nearly fulfilled for tho dato 
B.C. 220,000, but neither for B.C. 850,000 nor B.C. 210,000. 
Indeed at the latter date the winter solstice is almost exactly in 
the perihelion, and by the hypothesis tho midwinter heat 
would be 21° higher than now, instead of 38° lower. In the 
other case the solstice has the anomaly 1 26°*22/ by the approxi- 
mate reckoning. Tho rate of change is 139 o, 50 + 29*47= 
