298 ME. H. B. WOOBWAED ON A WELL-SINKIXa AT SWINDON. 



Sections, however, are very rare, so that but little can be said 

 about the southern range of the Kellawajs E-ock. The lowest parts 

 of the clay are perhaps too sandy for brick-making, and the ground 

 is as a rule too low to have been opened up in railway-cuttings. 



South of Cirencester, Prof. Hull observes that at the base of the 

 Oxford Clay " there occurs a thin series of yellowish sands and cal- 

 careous sandstone, highly fossiliferous, many of the fossils being 

 those of the Oxford Clay " *. Traces of the Kellaways Rock also 

 occur near Malmesbury. 



Further north the variable thickness and mineral character of the 

 " Kellaways Sands, Sandstones, and Clays," in their extension from 

 Wiltshire to Yorkshire, have been remarked on by Prof. Judd t and 

 Mr. HudlestonJ. 



The Combrash and the Porest-marble call for no particular 

 remarks ; many of the characteristic fossils occur, and in the latter 

 formation we find Terehratula coarctafa, which is suggestive of the 

 Bradford-Clay zone. 



4. We now come to the Saline Waters at Swindon, and to the 

 question of their source. 



The water from the Corallian beds, analyzed by Mr. F. W. Harris, 

 showed 144-2 grains of solid matter in the imperial gallon ; this 

 included 86 grains of sodium chloride, and 49 grains of sodium carbo- 

 nate. • The water from the Porest-marble, also analyzed by Mr. 

 Harris, showed 2131*85 grains in the imperial gallon, and this com- 

 prised 1824 grains of sodium chloride, 191 grains of calcium chloride, 

 88 grains of magnesium chloride, &c. The amount of sahne matter 

 was therefore nearly 15 times greater than in the upper waters. 

 Ko sodium carbonate was detected in the lower springs, and these, as 

 remarked by Mr. Harris, present some similarity to ordinary sea- 

 water. Thus the amount of sodium chloride was almost identical (1824 

 grains at Swindon, and 1964 grains in sea-water), and so also with 

 potassium chloride (16 grains at Swindon, 15 grains in sea-water). 

 On the other hand calcium chloride, of which 191 grains occur in 

 the Swindon water, is not noticed in sea-water ; while the latter 

 contains a much larger proportion of magnesium chloride, and also 

 considerable quantities of magnesium sulphate and calcium sulphate, 

 represented only by 1 grain of calcium sulphate in the Swindon 

 water. With this exception the absence of sulphates from the 

 Swindon water is noteworthy. 



The following analyses show the number of grains per imperial 

 gallon § : — 



■^ Geology of parts of Wilts and Gloucestershire (Mem. Geol. Survey), p. 19. 



t Geology of Rutland, &c. (Mem. Geol. Survey), p. 232. 



I Proe. Geol. Assoc, vol. iv. p. 410. 



§ Some of these analyses were published by Prof. Prestwich in a paper read 

 before the Ashmolean Society, 1876. See also C. E. De Eance, Trans. Man- 

 chester Geol. Soc, Dec. 1884. 



