254 MR. W. H. DINES ON THE VERTICAL TEMPERATURE 



that are left are not all available for the following reason. There are two sets of 

 twenty relating to the same day, which were sent up at Manchester (at hourly 

 intervals) for the purpose of investigating the daily temperature variation, and it is 

 obvious that if all these were used in forming an average value, that value would 1* 

 vitiated if any unusual conditions prevailed on the days in question. On other days, 

 too, notably on the date of the supposed passage of the earth through the tail of 

 H ALLEY'S comet, the individual ascents are crowded too closely together to render 

 them fit for the purpose. In cases where more than two ascents, have occurred 

 within eight hours the mean of such ascents has been formed and that mean taken as 

 equivalent to an individual ascent. 



The ascents have been made chiefly at Manchester, in Oxfordshire, and in Sussex, 

 but there are 14 from Scotland and 6 from Ireland. All meteorologists are indebted 

 to the liberality of Prof. SCHUSTER and to the Manchester University for the 145 

 balloons sent up there, of which 105 have been found. They are also indebted to 

 Mr. C. J. P. CAVE who, notwithstanding the obstacle of his geographical situation at 

 Petersfield, near to the Sussex coast-line, has made 42 successful ascents out of a 

 total of 70. Pyrton Hill, 16 miles S.E. of Oxford, is the official station of the 

 Meteorological Office. From it, and from Crinan, N.B., 125 balloons have been sent 

 up and 81 found. For the others we have to thank the Joint Committee of the 

 Koyal Meteorological Society and the British Association. 



The observations are fairly well distributed over the three years. The ascents have 

 been made (mostly about the time of sunset) on days previously appointed by the 

 International Aeronautical Commission, and thus it happens that three ascents are 

 often available for the same day, one from Manchester, one from Pyrton Hill, and one 

 from Ditcham Park, Petersfield. There is generally close agreement between the 

 results from Ditcham Park and Pyrton Hill, since the stations are only 40 miles 

 apart, hence the observations cannot be accounted independent, but in forming the 

 averages the two observations have the merit of helping to cancel out any chance 

 instrumental errors. 



It cannot be contended that 200 observations distributed more or less by chance 

 into 36 weeks spread over a period of three years will suffice to give good average 

 values for the temperature of the upper air, but it seems desirable that the figures 

 should be worked up and the best results that they afford obtained. It is also of 

 interest to see what agreement there is with the tables similarly obtained on the 

 Continent. 



Annual Temperature Variation. 



To obtain the annual temperature variation the observations have been sorted into 



ths, and the mean temperature for each month at each exact kilometre of height 



The crude result is given in Table I., but it is obvious that the values may 



noothed with advantage. Also, anyone working up these figures cannot fail to 



