74 REPORT — 1855. 



is lost to the rivers and springs by evaporation, or by being taken up by 

 vegetation. The physical and geological features of the country will produce 

 very varying results. The proportionate quantity of water which will flow 

 from steep mountain sides, consisting of impervious rocks, will be very dif- 

 ferent from that which will pass away from a gently undulating country well 

 clothed with vegetation. 



The first accurate observer on a large scale in this department appears to 

 have been the late ingenious Mr. Tliom of Rothesay, the constructor of the 

 Shavvs Water-Works, near Greenock. 



The following is the result of information which he gave some years ago 

 to the Institution of Civil Engineers on the rain which fell in 1826 and in 

 1828, the former year being the dryest year on record, and the latter, one in 

 which there fell more than the average amount of rain : — 



inches. 

 From the 1st April 1826 to 1st April 1827, the fall of rain in Bute was 4>5-4i 



Of which there found its way to the reservoirs 23-9 



Lost to the reservoir 21"5 



In 1828 the rain at Greenock Reservoir was 60 inches, of which there 

 flowed to the reservoir 4-1 inches, showing a loss by evaporation, vegetation, 

 absorption, &c., of 19 inches. Further observations by Mr. Thom led him 

 to the conclusion, that the loss bore a certain definite proportion to the rain- 

 fall ; and the late Mr. Stirrat of Paisley, also an accurate observer, viewed 

 the question in the same light ; their average results giving the loss at about 

 ^ths or Y^oths of the whole fall, when the annual amount was from 54 to 65 

 inches. This conclusion was no doubt correctly arrived at from the facts 

 before them, but it is obvious from a little reflection that this mode of calcu- 

 lation is inapplicable to other districts, where a much larger or a much 

 smaller quantity of rain might fall. For instance, the requirements of vege- 

 tation and the amount of evaporation are usually much less where a large 

 quantity of rain falls, while at the same time the ground is generally less 

 absorbent and the declivities greater, and it evidently follows that the 

 loss by evaporation and vegetation must be less under such circumstances 

 than in a rich level country, where the rain is not nearly so great. By as- 

 suming a certain definite proportion of the whole rain, the reverse would 

 appear to be the case. Take, by way of illustration, 100 inches in a sterile 

 mountainous country, the loss at ^ths would be 30 inches ; and take 30 

 inches again as the rain in a fertile level country, the loss at -^ths would be 

 but 9 inches, obviously inconsistent with the real facts of the case. The truth 

 appears to be, that the loss within certain limits is a tolerably constant quantity, 

 and that generally the greater the rain the less the deduction ought to be. 



The observations of Mr. Thom and Mr. Stirrat alluded to, give the annual 

 loss at from 18 to 23 inches per annum, out of rain-falls of 54 inches and 65 

 inches respectively. Measurements and observations in 1852 in the Gorbals 

 Water-Works district, closely adjoining those in which these observations 

 were made, and in which there is about the same amount of rain, show the loss 

 to have been but 12 inches out of 60. The average loss from several years' 

 observations at the Manchester Water- Works is about 12 inches per annum. 

 Mr. Hawksley's observations at the Liverpool New Water- Works, in 1847, 

 show a loss of 12|^ inches. 



Other observations scattered over the country show the loss to be ordi- 

 narily from 12 to 16 inches, and to a great extent to be irrespective of the rain 

 which falls. In determining, therefore, the probable quantity of water which 

 maybe collected from any district, other than one of an absorbent character, it 

 is necessary first to ascertain the fall of rain, and then, having due regard to 



