Vol. 55.] ox PALEOZOIC EADIOLAEIAX SOCKS IN If. S. WALES. 31 



present for assigning a greater antiquity to the Jenolau Cave 

 radiolarian rocks than to those at Tamworth. At all events, both 

 may very well fall within the limits of some part of the Devonian 

 Period, though on the whole it is probable, in view of the abundance 

 of large specimens of Pentamerus Kniglitii in the Jenolan Cave lime- 

 stone, that that rock and its associated radiolarian shales are older 

 than the radiolarian shales and limestones of Tamworth. 



With respect to the submarine tuff, Mr. G. W. Card has 

 pointed out to us that a number of thin slices of this rock prepared 

 under his supervision have exactly the aspect of a felsite-tuff, 

 numerous fragments of crjptocrystalline felsite being entangled in 

 the holocrystalline or microcrystalline groundmass. The latter is 

 composed of broken and corroded crystals of plagioclase, orthoclase, 

 quartz, and augite, with occasionally hornblende, and more rarely 

 sphene. Small crystals of iron-pyrites are numerous, and grains of 

 titaniferous iron sometimes occur ; small and large enclosures of the 

 radiolarian rocks abound. 



The thickness of the submarine tuff-beds varies from that of a 

 mere thread up to about 100 feet or more. At some depth from 

 the surface the colour of the tuff is greenish-grey, weathering to 

 yellowish-brown or light grey at the surface, and thus contrasting 

 strongly with the darker claystones. A chemical analysis of the 

 tuff will be found on p. 32. 



Reference has already been made to the occurrence of casts of 

 Le;pidodendron australe on three horizons in these tuffs (p. 27), and 

 also to the fact that the bed developed at Cleary's Hill and in the 

 spurs to the east and north (pp. 23, 24) has all the appearance of a 

 volcanic agglomerate, with large included fragments and pebbles. 



IV. Chemical Compositiois^. 



The following analyses (p. 32) of the radiolarian limestones, cherts, 

 shales, tuffs, and volcanic agglomerate were made by Mr. J". C. 

 H. Mingaye, F.C.S., Analyst and Assayer to the Geological Survey 

 of New South Wales, from samples collected by us. 



With reference to these analyses, we would comment on the 

 fact that although, so far as can be judged from the microscopic 

 examination, the radiolarian shales are almost as rich in radiolarian 

 remains as the black chert, the former contain only about 68 per 

 cent, of silica, while the latter contains 91*06. This points, in our 

 opinion, to the probability that the higher percentage of silica in 

 the chert is due, not to the silica which it has received from the 

 radiolarian tests, but rather to secondary silica derived from the 

 siliceous tuffs. This supposition, however, is not proved, as, 

 owing to the minuteness of the particles forming the base in 

 which the radiolarian tests are embedded, it is almost impossible 

 to determine whether that base was originally siliceous (formed 

 perhaps of comminuted fragments of radiolaria) or partly aluminous. 

 If, therefore, the cherts have been formed from the shales through 



