42 
Trunk Drainafjc. 
it is important to observe that, in these instances of fen drainasje, 
the substratum is known to be impervious ; and by the adoption 
of catch-water drains, intercepting the waters from adjacent 
hij;h(!r lands, the water that falls directly upon the land may be 
precisely calculated [not accurately, however, because of the 
soakage through the embankments, and the water let in during 
the summer] ; but with the physical peculiarities of the Test 
valley we have not only to provide for rain falling directly upon 
the land, but also for the springs, sock, and under-water, which 
rise to the surface, and penetrate the peaty soil from the sur- 
rounding chalk. Taking the average fall of rain' in the valley 
of the Test to be 30 inches, and the quantity of water finding its 
way into the valley from the several sources referred to, to be 
three times greater than that of the fens, we have taken 7 inches 
per acre, or 25,740 cubic feet, to be the maximum quantity of 
water it is necessary the outfall drains should be capable of 
dischai'ging per montli ; and this quantity is equal to 6-lOths 
of a cubic foot per acre per minute." We have pleasure in 
transcribing these extracts for the information of our readers, 
because they give a most admirable pattern of the way in which, 
by help of data derived not only from observations of one peculiar 
district, but also from kindred localities, and likewise from facts 
of science, the selection and construction of the drains are fixed 
and governed without waste or miscalculation ; are brought, 
indeed, under the simple formulas of hydraulic engineering. 
The report proposes to make all the outfall drains of nearly 
doul)le this theoretically-deduced capacity, and to follow out the 
improvement by a complete ramification of open and under 
drains, l)y which the pores of the soil may be kept open, pent-up 
waters released, and springs tapped, confined to, and discharged 
by, separate channels. To satisfy any persons who may object 
that a subsoil drainage of 4 feet will render dieir water-meadows 
too dry, it is suggested that " each outlet or junction of an 
under-drain into and with the outfall drains, may be provided 
with a self-acting trap and fixed Ijolt, or with a plug, by the 
application of which all water may be stayed in the drains until 
re-opened." An ingenious thouglit is then enunciated, which 
may be of general value when river flats have more extensively 
experienced a practical application of the purport of this essay. 
" A new method of summer watering the soil has presented 
itself to our minds during the investigation which may be found 
applicable ; a system of underground inlet drains or channels of 
j)ipes placed intermediate b(!tween the outlet drains, but at a 
higher level (say 12 inches), or 3 feet deep from the surface, may 
be devised so as to admit water flowing from the river, or from 
any other attainable source. The water flowing down those 
