518 On Draining. 



inches in diameter and three feet deep, there were inserted into it two 

 pipes turned downwards for the water to run off into bottles : the one pipe 

 was near the bottom of the vessel, the other was an inch from the top. 

 The vessel was filled up, for a few inches, with gravel and sand, and all 

 the rest with good fresh soil. Things being thus circumstanced, a regular 

 register has been kept of the quantity of rain-water that ran off from the 

 surface of the earth through the upper pipe (whilst that took place), and 

 also of the quantity of that which sank down through the three feet of 

 earth, and ran out through the lower pipe. A rain-gauge of the same dia- 

 meter was kept close by, to find the quantity of rain for any corresponding 

 time." 



You will notice that the general result of these experiments 

 accords pretty nearly with that of the Holmfield account ; and 

 yet it may be readily conceived that circumstances of situation 

 and stratification may often occasion as wide a difference in the 

 amount of filtration as is shown between Mr. Dickinson's and 

 Mr. Charnock's observations. 



On an examination of the details registered in the account 

 before us, it will be evident that the amount of filtration is not 

 exclusively dependent on the fall of rain ; but that a variety of 

 other causes combine to affect its proportion. For instance^ in 

 March, April, May, June, and July, of 1842, the fall of rain was 

 13*65 inches, and the filtration for the same period was only 2"05 

 inches ; whilst in April, 1846, there was 5*97 of rain and •2-99 of 

 filtration. Similar instances are also noticeable in Mr. Dickin- 

 son's details. From March to October, inclusive, of 1840, a fall 

 of 11*52 inches of rain is recorded, without any filtration ; but in 

 November, 1842, the rain was 5*77, with 5 inches of filtration. 

 Dr. Dalton's table also shows the same variations. The lesson, 

 therefore, derivable from these experiments, so far as regards fd- 

 tration by drains, is one rather of a speculative than of a definite 

 character ; for, although we are assured filtration must be secured, 

 we are left with a large and varying margin as to the proportion. 

 We must not, however, overlook the fact, that all the registered 

 details show occasionally an amount of filtration nearly equal to 

 the rain that falls, and therefore, in determining the size of pipe 

 to be used, the ready exit of this maximum quantity must be 

 provided for. 



Considering the Holmfield House observations as a whole, and 

 the result of each experiment in conjunction with the others, they 

 establish very conclusively this important fact — that drainage, to 

 be complete in its effects, must secure the greatest and most uni- 

 form amount of filtration of which the land is capable with a 

 given outlay ; for of course there is always a compensating balance 

 between cost and effect to be regarded. Seeing then that, even 

 in a comparatively open subsoil of gravel and calcareous sand, 

 this desirable proportion of the rain that falls only very occasion- 



