34 



KNOWLEDGE 



[Febbuaky 1, 1892. 



Curiously this law of error was overlooked by Vega in 

 his splendid reprint of Vlacq's "Logarithms" to 10 places. 

 Wishing to facilitate the correct calculation of intermediate 

 logarithms to 10 places, he gave a table of what he called 

 " second differences," which were to be added to the 

 results taken by means of proportional parts. But 

 omitting to take notice that these -'errors "or "second 

 differences " decrease inversely as the tupiKres of the 

 numbers, his table, throughout the greater part of it, 

 makes the correetions greater than they should jiroperly 

 have been. T. B. Baerett. 



THE CAUSES OF THE ICE AGE. 

 2o the Editor of Knowledge. 



Sir, — I am sure that my friend Sir Eobert Ball would 

 not have made the charges which he has made against 

 Sir John Herschel and Dr. Croll without good reason, 

 though I am not aware of the passages on which it is 

 based. But his article seems to imply that the general 

 reasoning of both authors rests on this erroneous assump- 

 tion, which is certainly not the case. Both assume (as 

 indeed every writer on the subject must do) that each 

 hemisphere receives more heat during its summer than 

 during its winter. This is clearly implied, for instance, 

 in 368(( of Herschel's Outlines ; but neither he nor 

 Dr. Croll, I think, have anywhere attempted to compute 

 the relative quantities of heat which we receive during our 

 summer and our winter. This, however, was done in a 

 very complete manner by Dr. Haughton in The Trans- 

 actions of the Fioijal Irish Acailenuj for 1881. His figures 

 for the entire hemisphere agree pretty well with Sir Eobert 

 Ball's, but they vary with the latitude of the place. 



Sir Eobert Ball is probably correct in setting down the 

 effects of the solar heat as equivalent to a mean elevation 

 of temperature of more than 300^ F. This elevation, 

 if distributed in the ratio of 63 to 37, would give a 

 mean elevation of 378° F. in summer and ^22° F. in 

 winter, with an annual range of 156° F. Of course, 

 nothing like this annual range exists, the reason 

 being that modifying causes (especially air-currents 

 and ocean-currents) come into play. Now unless it 

 can be shown that these modifying causes do not come 

 into play in the case of the secular variation arising from 

 changes in the eccentricity of the earth's orbit, the amount 

 of the secular variation must be proportionally reduced. 

 This, accordingly, is Dr. Haughton's conclusion. For 

 lat. 50° he makes the annual variation of temperature 

 41-2° F., and the secular variation only 3-685° F., more 

 than one-half of which will be represented by a summer 

 elevation of temperature instead of a winter depression. 

 Sir Eobert Ball's reference to the large variation in the 

 interior of continents is, I may remark, not much to the 

 purpose, since it is not in the interior of continents but in 

 islands and continental countries not far from the shore 

 that the traces of glaciation are most evident ; and in 

 Siberia, for instance, where this great annual range of 

 temperature exists, there is, generally speaking, no 

 permanent snow-cap. 



But behind all this lies the further question, Is such 

 an unequal distribution of temperature favourable to the 

 production of a permanent snow-cap •? Dr. CroU's argu- 

 ments on this point seem to me imsatisfactory. Sir 

 Eobert Ball (as far as I am aware) has not argued it 

 at all. Sujjpose, as an extreme case, that the winter 

 37 per cent, of heat was distributed over 364 days 

 in the year, and that we then received the remaining 

 63 per cent, on the 365th day. Would this 63 per 

 cent, of the total heat be incapable of melting the , 



snow-cap produced by the 364 cold days ? I think 

 not. in our latitudes 63 per cent, of the total annual 

 heat would suffice to melt at least 50 feet of solid ice. 

 It would no doubt have to raise the ice from a low tem- 

 perature to freezing point before the melting commenced ; 

 but then, without a great increase in our rainfall or 

 snowfall, it would not have 5 feet of ice to melt instead 

 of 50. The snow might be more than 5 feet deep, 

 but it would not be equivalent in amount to more than 

 5 feet of ice. In fact, a given quantity of heat would, it 

 seems to me, produce its maximum melting effect when 

 brought to bear in the shortest time. Suppose the heat 

 received at a certain place is sufficient to maintain a 

 constant temperature of 30° F. If the distribution is 

 uniform, and the locaUty is sufficiently supplied with 

 moisture, there wiU be a permanent snow-cap ; but, 

 unless the snowfall is very large, could I not melt the 

 snow-cap once a year if allowed to distribute the heat as 

 unequally as I chose ? 



This seems to me to be the most important question to 

 be decided. As to Dr. CroU's theory of the diversion of 

 the Gulf Stream, the fact that glaciation extends much 

 farther to the south on the American than on the European 

 side of the Atlantic affords pretty strong evidence that the 

 Gulf Stream was not diverted during the Glacial Period. 



Dr. Haughton takes occasion to point out that the 

 secular range of temperature could not account for the 

 miocene coal-beds of North GrinneU Land ; but the 

 possibility, or impossibility, of accoimting for this and 

 other evidences of a former mild chmate in the Arctic 

 regions is regarded by Dr. CroU as a crucial test of his 

 theory. I remain, yours faithfully, 



Dublin, Jan. 13th, 1892. . W. H. S. Monck. 



[One of the difficulties to my mind in accepting the 

 explanation of the cause of the Glacial Epoch suggested 

 by Dr. Croll, and adopted by Sir Eobert Ball, is that, if 

 we could rely on such calculations, Mars should be per- 

 manently covered with snow, whereas we see its white 

 polar caps was and wane with the Martian summer and 

 winter, proving, it seems to me, that there must be very 

 potent modifying influences (such as the existence of a 

 dense atmosphere) which upset the assumption that the 

 mean temperatitre of a planet's sm-face may be assumed 

 to vary inversely as the square of the planet's distance from 

 the sun. Our atmosphere is probably not constant in 

 quantity through geologic periods. Its gases are, we know, 

 continually being absorbed by one set of chemical actions, 

 and set fi'ee by others. The one set of actions may exactly 

 balance the other, but if not, the atmosphere must be 

 growing or decreasing in amount, and this woidd pro- 

 foundly modify the sttrface temperature. So also would 

 the elevation of an ocean bed which turned aside a 

 tropical cm-rent such as the Gulf Stream. — A. C. Eanyakd.] 



Dr. Max W^olf is continuing his photographic explora- 

 tions of the Milky Way. In a letter received three or four 

 weeks ago he informs us that several of his negatives show 

 the tracks of meteors which have been observed to pass 

 across the region photographed during the exposure. This 

 gives the hope that in the future rich meteor showers may 

 be observed photographically and the area of the heavens 

 from which they radiate more exactly mapped than hitherto. 



* ♦ » 



Prof. W. W. Payxe, of Northfield, Minnesota, and Prof, 

 George E. Hale, of Chicago, have published the first 

 number of a new astronomical journal, which gives pro- 

 mise of occupying a very high place in astronomical litera- 

 ture. Prof. Hale states that he has undertaken the work 

 at the instigation of his fi'iend. Prof. C. A. Young, who 



