ON NINE BALLOON ASCENTS IN 1863 AND 1864'. 285 



The mimbers in the following Table differ very much from those in Table VI. 

 in the Reports for the year 1862 and ISG'6 ; in these the largest numbers were 

 those at the bottom of the column, and the smaller at higher elevations, 

 and a decrease of temperature with elevation was shown Avithout exception. 



In the following Table there are instances of departure from both those in- 

 dications, and other particulars which present aU the above numbers to be 

 combined with previous results. 



In the first remarks on Table III. at page 2G6, the decrease of temperature 

 as observed in the first 200 feet was no less than S° ; by the adopted curves 

 passing nearly through the observed temperatures, it gives 2^° decline in 

 each 100 feet near the earth ; these results seem doubtful ; they differ so much 

 from all others, that it seems likely that the readings were affected by the 

 presence of many persons near the car of the balloon before starting, or that 

 I have read the instruments Avrongly by 5° before leaving the earth. 



On January 12 (the only winter ascent in the series) the numbers are for 

 the first time affected by the sign—, showing an increase of temperature 

 with increase of elevation, and the numbers near the earth are smaller than 

 those at higher elevations. 



On April 6 the numbers are also anomalous as compared with those 

 previously obtained ; near the earth there was no change, and then a large 

 change, and higher still some numbers are affected with the — sign. 



On June 13, on descending at the time of sunset, it wUl be seen that there 

 was scarcely any change of temperature for 1500 feet. 



On June 20, on descending a little before sunset, the change was very 

 small, and very different from corresponding changes on the ascent an hour 

 before. 



On June 27 there seemed scarcely any change in temperature up to 

 3000 feet, at readings taken after sunset, and till it was too dark to read the 

 instruments. 



The results on none of these days can be used in deducing general laws. 

 The endeavour has been in the past year to take obseiwations at times in 

 the day and times in the year at which no observations had previously been 

 made, resulting in these very different results. 



The only days this year available for general combination are September 29, 

 from observations made between 8^ a.m. and 10*" a.m., with a chiefly cloudy 

 sky ; and the numbers in column 28 show the mean values at the different 

 elevations, and October 9 and August 29 for clear skies, and these results 

 are shown in column 30. 



The numbers in column 32 show the general mean from aU the observa- 

 tions with cloudy skies, as based upon the number of experiments as shown 

 in column 33 at each elevation, and these vary from 19 to 29. 



The numbers in column 35 show the results for clear or nearly clear skies, 

 as based on the number of experiments as shown in the column 36, varying 

 in number from 8 to 12. 



In column 34 the space in feet is shown for an increase of 1° with cloudy 

 skies, varying from 167 feet near the earth to 334 at heights exceeding 3000 

 feet. 



In column 37 the same results are shown for clear skies, being 143 feet 

 near the earth, gradually increasing to 334 feet at heights exceeding 1600 

 feet. 



