368 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [JimO 19, 



Quorndon ; fine-grained, slightly vesicular, with a little carbonate 

 of lime disseminated, and in cavities. 



For the five last named, as well as for many characteristic speci- 

 mens of the Malvern dykes, I am indebted to Dr. HolL 



General Conclusions. 



1. In the intrusive trap-rocks, the ratio of the oxygen of the 

 silica to that of all the bases taken together varies progressively 

 from 5 : 4 to nearly 2:1. The middle term, representing the mean 

 composition of the trap-rocks, is 3 : 2, which is the proportion assigned 

 by Bunsen to '^ augitic " rock of normal composition. It should, 

 however, be observed that the bases are not entirely combined with 

 the silica, but are, in various degrees in difi^erent rocks, united with 

 other acids, or deposited as free oxides in the capillary interstices of 

 the rock. This is easily proved by digesting the pounded rock in 

 diluted acetic acid, which dissolves out a portion of the bases with- 

 out decomposing the siUcates. There are also sulphates, bisulphu- 

 rets, fluorides, arseniates, and carbonates. Sulphate of baryta 

 occurs in cavities and fissures, and as a cementing substance ; it has 

 clearly been precipitated from a state of solution, and, being itself 

 insoluble, it must have been formed by the decomposition of carbo- 

 nate or silicate of baryta by soluble sulphates. Bisulphurets abound, 

 generally as pyrites ; arseniates south-west of the Midsummer Hill 

 and in the Eagged Stone ; fluorides in the trap-dykes and lava-beds, 

 most abundantly in those of recent formation, as the bosses in the 

 field near Fowlett's farm. Carbonates are constituents of some of the 

 rocks at the south end of the chain. Protoxide, peroxide, and mag- 

 netic oxide of iron are disseminated throughout the substance of the 

 intrusive rocks in various proportions. The proportion of the oxygen 

 of the silica to that of the bases actually combined tuith it, is there- 

 fore much nearer that of a bisilicate than it would aj)pear to be from 

 the above analyses. 



2. The relative proportions of the several bases vary considerably 

 in diff'erent rocks, and often characterize particular localities. Mag- 

 nesia abounds in the lava-beds and dykes, and in the volcanic ash of 

 the Hollybush-sandstone and Black- shale periods, in one or two 

 dykes of the Herefordshire Beacon, and in the Brock HiU. dyke. 

 The other traps, including the Bartestree dyke, though of the same 

 age as that at Brock Hill, have more lime than magnesia. Lime 

 and iron both abound in some of the dykes of the Herefordshire 

 Beacon, and at Glynrhonwy ; lime only in others — in the dyke at the 

 south end of Midsummer Hill, and at Bartestree ; alkalies in the 

 Eagged Stone and at Quorndon, and in some of the Herefordshire 

 Beacon. I have determined the alkalies as difference, but I ascer- 

 tained, from separate experiments, that there was more potash than 

 soda in the eruptive rocks associated with the Hollybush Sandstones, 

 and that in most of the other rocks examined for the determination 

 of the relative amounts of the two alkalies, soda was the most 

 abundant. 



3. The chemical composition of the eruptive rocks does not vary 



