Bridge sinks before reaching the swallow, and passes under the bed of the Alyn.

I then had the Garth Brook dammed and all turned down the swallow, and gauged the reissue at rate of 900,000 g. p. d. About 146 yards below the junction, what remains of the water brought down by the Garth Brook and reissue disappears down another swallow in the bed of the Alyn, again leaving the bed dry. I gauged it 40 yards above this swallow at rate of 644,000 g. p. d., showing a loss of about 256,000 g. p. d. in 114 yards as it flowed. For nearly mile below this Pont Newydd swallow the bed was dry, after which there was water which rapidly gathered volume all the way to Rhyd-y-mwyn. There was now due:

mile

Atmile below the Pont Newydd swallow, below the reissue and opposite Trumian, the Alyn gauged 950,000, leaving 4,450,000 g. p. d. still due. Three-quarters of a mile lower down and nearly mile above Rhyd-y-mwyn, where it leaves the limestone, the Alyn gauged at rate of 5,687,000 g. p. d., showing all the lost waters recovered, with others not investigated.

Early in 1887 the Bristol Consumers' Water Bill afforded the opportunity of surveying the whole area in Monmouthshire affected by the heavy pumping from the Severn Tunnel. Early in April of that year they were pumping a total of over 29 million gallons a day out of 5 shafts, the marsh shaft, the 5 mile 4 chains shaft, the 12 feet shaft, the iron shaft, and the 29 feet or Sudbrook shaft, at which point all the waters met with in the big spring

in the mountain limestone are raised to the surface and turned into the sea. Beneath the main tunnel an adit was driven. At 4 miles 50 chains from the south end, the mountain limestone rose into the adit, and after rising into the tunnel rolled over and disappeared again at the 5 miles 4 chains shaft. This hog's back of limestone was on the line of an anticlinal that was traceable across country, slightly curving northwards to Shire Newton. It acts as a subterranean watershed line, to the east of which the effects of the pumping have not been felt. In this limestone, which rose 15 feet too high for the economy of the tunnel, the fissure yielding the big spring was cut. The permanent pumping began at Sudbrook shaft early in 1885, drying up several large springs from the mountain limestone at the fringe of new red strata skirting the marshes. Round these springs had clustered the villages of Portskewet, Ifton, Rogiet, Llanfihangel, and Undy. In the limestone itself the effects are traceable for 5 miles to the west and northwest; due north, from the absence of wells, it was not possible to fix the limit, but a spring called Broadwell, a little over 2 miles N. of Sudbrook, which gauged 272 g. p. m. April 9, 1887, 660 yards below the spring, was unaffected, as were two pumps in wells at Ifton Hill, just over a mile E. of N., and at Hay's Gate well in the limestone a little further off in the same direction. Broadwell flows S. W. in the Nedden basin, and sinks partly into the limestone before it reaches Crick. The limestone is afterwards exhausted by the Sudbrook pumping.

A well in the limestone at Five Lanes, 4 miles N.W. from Sudbrook, ground level 62 feet, in the bottom of the valley, and 33 feet deep, which had 30 feet of water before the pumping began, was drained entirely. At Carrow Hill, 4 miles N.W. by W. from Sudbrook, a

well in the limestone, 32 feet deep, with 9 feet of water, went dry the very day after the pumps were started. It was then deepened 9 feet, and was dry in April, 1887, being then 41 feet deep, which gives a depression of at least 18 feet at a distance of 42 miles. At Ifton, 23 miles W. by N. from Sudbrook, a well was sunk in the bed of the dried spring 30 feet, and the depth to the water was 15 feet April 11, 1887. The parapet is 2 feet above the original water level of the spring, so that the depression at that point was 13 feet. At Rogiet a shaft was sunk in the same way, and the depression there was 12 feet. Llanvaches Brook flows in a small basin 0.52 square mile of old red sandstone, and was all spring water April 11, 1887. Directly it gets to the limestone most of it sinks near Red House, 8 g. p. m. still flowing at a farm called Cayo on that date. The bulk of Llanvaches Brook, which sinks at Red House, reissues at Fountain Spring, or Rockwell, from a cavern of limestone into Penhow Brook on the east side, and gauged 69 cubic feet, or 430 g. p. m. April 11, 1887. Just below its junction with Penhow Brook was another gauge which showed 87.42 cubic feet, or 546 g. p. m., thus leaving only 116 g. p. m. coming down Penhow Brook from the remaining half-mile of limestone, and an area of 175 square mile of old red sandstone above. In dry weather the bulk of Penhow Brook comes from Fountain Spring, which is nearly 5 miles N.W. by N. from Sudbrook Shaft. Just below St. Bride's, a mile lower down Penhow Brook, and 5 miles from Sudbrook, a gauge showed only 63 cubic feet or 4025 g. p. m., thus showing a loss of 1445 g. p. m. by Penhow Brook in flowing that distance April 11, 1887. The 5-mile circle on the west fairly represents the limit of traceable effect on the limestone water, of the pumping at Sudbrook shaft, as far north as the Caerwent and Newport Road.

The Llanvair Brook, which comes off the old red sandstone through Llanvair Discoed, was not gauged. It becomes dispersed over the meadows on the lower limestone shales below the village, and all sinks into the limestone.

The Castroggy Brook gauged at Llanmellin gauge, April 11, 1887, gave 160 cf. per min., or 1,435,392 g. p. d. Thence it flows down a concrete channel to The Whirlpools, originally springs, but now all dry, except one which acts as a swallow hole, and carries down all that is left of the brook to join the big spring at Sudbrook Shaft.

It was impossible to gauge all the waters which flow into the limestone basin that yields the big spring. Those that were taken on April 9 and April 11 give only:

Difference between St. Bride's and

9. p. d.

leaving 10 millions g. p. d. not gauged.

The total area contributory to the big spring is made

The lowest return for the big spring at this date was 12 million g. p. d. Taking the area at 21 square miles, a percolation of 13 inches per annum gives 10,828,998 g. p. d., to which must be added about 2 million gallons per day from the Newport area of O. R. S., being water which they do not take, making in all about 13 million g. p. d., or just sufficient to account for the big spring, which ranges from 12 to 15 millions. The area of new red beds fringing the marshes is 3.25 square miles. The water from this stratum is separately pumped in the 5 mile 4 chain shaft, where the pumps are capable of raising 10 million g. p. d. The marsh shaft takes rain water flowing towards the tunnel, and the iron shaft and 12 foot shaft, close to the sea, brackish water.

GUERNSEY.

In February, 1888, I was instructed to survey the island, and report to the concessionaires of the long proposed Guernsey Waterworks as to the best site or sites for a well or wells. The southern half of the island of Guernsey consists of an elevated tableland with cliffs on the south coast, and a slope towards the N.W. and N. The top of this plateau is flat, and consists of an earthy or gritty residue (there called "gravel ") after the disintegration of the subjacent rocks, which are mainly gneiss of a coarse description, slightly foliated, lying vertical or at high angles, striking N. and S. or W. of N. and E. of S. It is much intersected by feldspathic trap dykes. The disintegrated part descends to the low ground with the fall of the gneiss ground towards the N.W., in which slope is a series of somewhat shallow valleys. The principal of these empty into Perelle Bay and Vazon Bay.