194 report — 1859. 



viz. about the 3rd day before first quarter (or the 4.-th or 5th day of the moon's 

 age), and at three other noticeable points of the curve ; these, as I have 

 already stated, are (1) shortly before and (2) shortly after full moon, and (3) 

 immediately after last quarter. 



In the colder half of the lunation the temperature rises at full moon, and 

 shortly before last quarter. 



2. It is not, however, only upon an average of a long series of yearly means 

 that the proof of lunar influence depends for its establishment. All the more 

 remarkable deviations from the calculated mean temperatures of the day, for 

 which the past year has been distinguished, have followed the model curve in 

 a more or less significant manner, at the seasons of excessive heat or cold above 

 alluded to. I have elsewhere shown that this was the case in November 

 1858*, and in January, March, and April in the present year j- ; and it has 

 since proved to be so in September and October, and thus the greatest amount 

 of heat on the average of 12 lunations in 1859 displays itself according to the 

 rule above indicated in the first half of the curve, the greatest amount of 

 cold in the second half (see fig. 2). The mean of the means of the several 

 columns is 51°*1 ; the mean of the first 14 columns 51°"9, of the remaining 14 

 columns 50 o, 2. The table of mean temperatures from which this curve was 

 formed is appended, in order to afford those who may wish to examine more 

 minutely the nature of the influence exerted in separate lunations, an oppor- 

 tunity of doing so. It will also serve to illustrate the method which was 

 adopted in arranging the observations in the several lunar tables from which 

 the curves have been obtained. 



3. The popular belief in a tendency in the weather to " clear up," or the con- 

 trary, at certain periods of themoon'sage — a notionwhich my own observations 

 appeared to confirm — -joined with a strong impression that these seasons 

 would be found to synchronize more or less closely, according to the time of 

 year, with the periods of greatest cold or heat in the lunation, led to the con- 

 clusion that the rise or fall in the curves of temperature must be due to the 

 action of terrestrial radiation, as a secondary cause ; and that the rise in tem- 

 perature at other periods of the lunation might also possibly be attributed to 

 the opposite state of the atmosphere when radiation upwards to the sky is 

 stopped, more particularly in winter, by the presence of low or thick masses 

 of cloud. This view has been much strengthened during the past year by 

 results which were obtained from an examination of the bi-horary observa- 

 tions of cloud taken day and night continuously for seven years (1840-47) 



* See Phil. Mag. for March 1859. 



f Whilst, according to the calculated average, the mean temperature at the end of March 

 and beginning of April ought, in each case, to rise above and fall below the mean of the 

 month, in 1859 this was exactly reversed. A very cold period occurred at the end of March, 

 the mean temperature (on the 31st) being 9° - 4 below the average of that day of the month 

 for forty-three years, as determined by Mr. Glaisher. But the 31st of March was also the 

 third day before new moon, and the mean temperature of that day of the lunation in March 

 for the same number of years falls below the mean temperature of the lunar curve. So also 

 in April, the mean temperature of the 7th day was 17°"5 in excess of the mean temperature 

 of that day for forty-three years at Greenwich, and the mean temperature of the 15th day 

 was 8°-3 below the average. Here the 7th day of April fell on the day of maximum tempe- 

 rature for the lunation in April or the first octant, and the 15th day of April was the second 

 day before full moon, which is within the cold period which precedes that phase of the moon. 

 The minimum temperature at the Toronto Observatory also in January 1859, which was 

 — 26°-5 on the 10th day, rose on the 13th to 36°0. At Greenwich a similar rise took place 

 from the 9th (or 3rd day before first quarter) to the 12th (or day of first quarter). On the 

 former day the minimum temperature was 28° - 5, on the latter 41°"2, and the mean tempe- 

 ratures 33° - 6 and 45°*0. There appeared to be a considerable development of electricity at 

 all the periods of low mean temperature. (From a communication made to the London Me- 

 teorological Society, and reported in the ' Athenaeum.') 



