456 EEPORT— 1891. 



printed in the ' Jonrnal ' of the Scottish Meteorological Society for 1888. 

 The observations of Mr. W. Watts on behalf of the Manchester Geo- 

 logical Society at the Piethorn and Deushaw Reservoirs have been 

 printed in full in the ' Transactions ' of that Society. Mr. Ashvs^orth's 

 observations on the Cowm and Spring Mill Reservoirs have also been 

 published in a summarised form with detailed curves in the ' Proceed- 

 ings ' of the Rochdale Literary and Scientific Society. 



In addition to the observations specially made or discussed for the 

 Committee, many papers are to be found in the ' Proceedings ' of the 

 Royal Society of Edinburgh, the publications of the Meteorological 

 Societies, in the annual reports of the Fishery Board for Scotland from 

 1887 onward, and in the ' Scottish Geographical Magazine.' A full 

 discussion by Dr. H. R. Mill of very detailed temperature observations 

 made hy the staff of the Scottish Marine Station on the Clyde Sea area 

 is nearly completed, and will be presented to the Royal Society of 

 Edinburgh. 



The best method of publishing the results of the observations dealt 

 with in this report has been carefully considered. There are peculiar 

 difficulties in dealing with a mass of data compiled by observers, some of 

 ■whom are skilled and others uninstructed in their work, especially when 

 — as in this case — the observations are taken with regularity in few cases 

 and often at different hours. The record of actual observations will be 

 preserved by the secretary of the Committee, who will be glad at any 

 time to place it at the disposal of any one interested in this branch of 

 meteorology. Weekly means have been calculated for each of the 

 stations where reasonably regular observations were made, and these are 

 published in the form of curves in this report. The curve-form was 

 chosen in preference to printing the figures, on account of the much 

 more vivid impression conveyed by inspection as to the relative air- and 

 water-temperatures and their seasonal variations. The unit of the curve 

 is the weekly mean, as in many cases on one or more days of the week 

 the observations were omitted, and a curve of actual daily readings 

 would present a very broken appearance, besides thi'owing into undue 

 prominence purely temporary variations. The monthly means are given 

 in tabular form. The monthly means are usually calculated as the average 

 of the four or five weekly periods comprised in the month ; but in the 

 case of very regular observations the monthly mean is given as the 

 average of the daily readings. Compai'ison of the two methods shows 

 practically identical results in the case of regular observations, and where 

 the observations are irregular the method adopted for general use 

 obviously gives more satisfactory results. On account of the much 

 greater difficulty of measuring the temperature of air, less reliance must 

 be placed on that pai't of the work than on the temperature of the water, 

 which is easily found in a readily comparable manner. 



For the drawing of most of these curves and the calculation of the 

 means from them the Committee has to acknowledge the assistance of 

 Mr. John Gunu, F.R.S.G.S. 



In several instances, when the observations were carried on by skilled 

 meteorologists in conjunction with the routine of a meteorological 

 station, the observers accompanied their records with a short summary 

 and discussion, in which the most striking relations of temperature were 

 pointed out. These statements, either completely or in abstract, are 

 embodied in this report. 



