718 GREAT SERPENTINE BELT OF NEW SOUTH WALES, iii., 



only 1 mm., across, but sometimes increasing up to a yard or more 

 in width. The rocks are sometimes richly radiolarian, and at other 

 times quite free from these organisms. The only apparent distinc- 

 tion is in the size of the constituent particles, for the radiolaria 

 do not usually occur in rocks in which the grains of sediment, 

 quartz, etc., are more than about one-third of the diameter of a 

 radiolarian test, and are best preserved in rocks of the finest grain- 

 size. This rule does not hold, however, for the tuffaceous bands, 

 and abundant radiolaria may be present in association with quite 

 large felspar-grains, as will be seen in Plate xxix., Fig. 18, from a 

 rock [M.B., 71], occurring one mile east of Barraba. Nor is it 

 true that the radiolaria are most abundant in the fine-grained 

 rocks ; the well preserved radiolaria of the exceedingly fine-grained 

 rock [M.B., — ], occurring near the limestone on Hall's Creek, 15 

 miles south of Bingara, are not nearly so abundant as in the more 

 coarsely-grained rocks of Barraba and Upper Manilla, or as in the 

 rock, N.T., 410, from near Nundle, shown in Plate xxix., Fig. 17. 

 which exhibits the effect of crushing. This rock is very rich in car- 

 bonaceous matter. 



The breccias and agglomerates occur in narrow bands here and 

 there, but need no special description, being very similar to the 

 Baldwin agglomerates, but are not so coarsely grained. Locally 

 also, the presence of pebbly and sandy bands has been noted 

 among the Barraba rocks. 



The tuffs are remarkable for the sharply bounded character of 

 their grains and in their acidity, for they are albite-oligoclases 

 with a refractive index distinctly lower than that of Canada bal- 

 sam. In some rocks, however, the plagioclase is a more basic oli- 

 goclase. In some specimens, a few grains of orthoclase occur. 

 The felspathic tuff is usually a creamy-coloured rock, and some- 

 times it may contain plant-stems. Occasionally, there are very 

 fresh, hard layers which, in hand-specimen, are most difficult to 

 tell from true igneous felsite, and the intrusions would be mapped 

 as sills, were it not for the absence of contact-effects. Such occur- 

 rences are found on Hawkin's Creek, near the Horton Road, west 

 of Barraba, in the hills east of Upper Manilla, and, again, on the 



