436 Mr. R. H. Scott on the [June 18, 



extent affected by certain peculiarities found to be very commonly pre- 

 sent in such photographic curves, and o which no satisfactory explana- 

 tion has as yet been discovered. The most important of these is what is 

 termed by us " bagging," the result of which is that the base or fiducial 

 line of the curve is no longer a straight line, but exhibits a certain de- 

 gree of curvature, so that the difficulty of determining the hourly or 

 other values by means of an engraved scale, bearing parallel straight 

 lines, is very considerable. 



At the suggestion of Mr. Francis Galton, the Meteorological Com- 

 mittee gave instructions that measurements should be made of the curves 

 by means of the instrument called Amsler's Planimeter, of which a full 

 description, by Mr. F. J. Bramwell, F.R.S.,*is printed in the ' Report of 

 the British Association ' for 1872. The object of this invention is de- 

 fined, in the paper quoted, to be " that the area of any figure, however 

 irregular, can be recorded in definite standard units of measurement by 

 the mere passage of a tracer along the perimeter of that figure." 



It is perfectly obvious that the measurement of the area of the curve, 

 if it can be executed with sufficient accuracy, must give a far more satis- 

 factory mode of ascertaining the value of the mean ordinate of the curve 

 than the calculation of the average of any number of measured individual 

 ordinates, while the economy of time ensured by the use of the planimeter 

 forms a most important recommendation for its use. 



The mode of employing the instrument is as follows : The entire 

 perimeter of the curve, down to the base-line, is measured, and the value 

 noted. Then using the same base-line, a rectangle of known height, in 

 units of the scale of the curve, is next measured in the same way, and 

 the value noted again. 



The ratio of these two values is the mean value of the ordinate of the 

 curve, or the mean pressure or temperature for the interval embrace'd by 

 the curve. 



It may be remarked that I have learnt within the last few days that 

 the present occasion is not the first on which a planimeter has been used 

 for the deduction of meteorological means. Mons. van Rysselberghe, 

 Professor at the School of Navigation, Ostend, has employed it in con- 

 nexion with his new electrical Meteorograph. 



The subjoined Table shows for a period of eight months the means of 

 temperature for Kew Observatory obtained by the planimeter, as well as 

 those yielded by the old method, both for daily and for five-day means. 

 It will be seen that the difference in 242 determinations of daily means 

 only amounted to 0'5 on six occasions, and to 0-6 in one instance ; while 

 out of 49 cases of five-day means the greatest difference was only 0< 4, 

 and this was only once attained. 



At the end of the Table the column headed " "Wr. Eep. Plates " gives 

 the values obtained by measurement of the plates published in the 

 1 Quarterly Weather Report ' for the period embraced by the measure- 



