124 PROCEEDINGS OP SECTIOM C. 



Thus a number of observations should give the value of U 

 correct to much less than one per cent. 



Again, suppose fi/'t=\, ct==cd—-01 and /(=2 ; 

 the substitution in equation 3 A gives 



0^/;,= — 1X-01 + -01 ; 



i.e., cfi/^i^ — -005+ -01, 

 and this gives a maximum value of .015 if r7 and td are of opposite 

 sign. This is an error of 1 "5 per cent., and a number or observations 

 would reduce the probable error to much less than one per cent. 

 As ct and rd are measured in millimetres this indicates that 



(1) For minerals with relatively low refractive indices a 



thickness of material not much less than 1*4 m.m. 

 should be used to give values correct to one per cent ; 



(2) For minerals with relatively high refractive indices a 



thickness of material not much less than 2. m.m. should 

 be used to achieve a similar result. 



Mr. Sorbyi states that if the thickness be from ^th to J inches 

 " the errors ought to be limited to the third place of decimals." 



Most that is contained in the foregoing may be represented 

 graphically as follow^ ;-^ 



If the curves /ii=t/{t~^d) be plotted for different values of /, 

 making ^t and d the variables, a series of rectangular hyperbolas is 

 obtained, all passing through the point 0,1 and all being asymptotic 

 to the line d=t for the particular value of t belonging to it. 



The curves for ^=1, 2, 3, 4 and 5 m.m. are represented in the 

 accompanying diagram. From these it is apparent that as /t in- 

 creases the curves become flatter, bending over to become asymp- 

 totic to the lines ^=/. Hence on any curve a small change cd 

 makes a greater change in the value of m when u is large than 

 when /( is smaller. If the ratio r/(//< be tested it will be found to 

 increase continuously with /*, and this corresponds to statement 

 (1) on page 121. 



Again, as t increases the curves rise more steeply, which means 

 that for any given value of /( the change t/ii, and hence the ratio 

 f>fi//ii, is smaller for thick than for thinner plates of material, and 

 this corresponds with statement (2) on page 122. 



G.—NOTES ON THE BRISBANE SCHISTS. 

 By R. A. WEARNE, B.A. 



Previous work on the subject. — A paper was read by Dr. H. I. 

 Jensen at the Brisbane (1909) meeting of the Association upon 

 " The Metamorphic Rocks of S.E. Queensland." He thus refers to 

 the Brisbane Schists : — 



" The Brisbane Schists, also provisionally marked Gympie by our 

 Geological Department, are so crushed, folded, foliated and 

 faulted, that they must be assigned to the Middle Zone, and 

 consequently they are likely to be older than the true Gympie. 

 \\'hat age we cannot say, but possibly Pre-Devonian." 



1 Loc. cii. 



