160 Transactions of the Royal Society of South Africa. 
Table 2 gives corresponding particulars for macles of prevailingly 
rliombic-dodecahedral character. 
From Tables 1 and 2 we gather that, on the whole, the smaller octahedral 
macles appear to have a slightly greater spread than the larger ones, the 
opposite being the case for the dodecahedral ones. Measurements of a very 
much larger number of specimens, however, would be required to definitely 
prove that it is so. What is clear is that the octahedral types have a larger 
spread than the dodecahedral ; for the 213 macles of Table 1, with an 
average edge of 6*5 mm., have an average spread of 2'65, whereas the 
179 macles of Table 2, with an average edge of 6 mm., have an average 
spread of 2*30. This is partly (but only partly) to be accounted for by 
the fact that the thicknesses in Table 1 are measured from face to face, 
whereas many of those of Table 2 had to be taken between two opposite 
coigns. 
Table 3. — Number of Octahedral Macles of Given Spread. 
E, T under 2. 
2 to 2-99. 
3 to 3-99. 
4 to 4*99. 
5 to 5-99. 
6 upwards. 
Koffyfontein 
6 
3 
4 
Jau^ersfontein 
4 
5 
Dutoitspan . 
4 
10 
5 
Bultfontein 
17 
52 
28 
6 
1 
1 
Wesselton . 
]2 
36 
17 
2 
Totals . 
43 
106 
54 
8 
1 
1 
For the whole 392 macles measured the average edge was 6*3 mm. and 
the spread 2 49. Whence it would appear that the average thickness 
of a made is very closely one-half that of the regular octahedron standing 
on an equal base. The question now is, Has this ratio anything more than a 
chance significance ? To test this query we must determine the actual 
number of specimens of given ratios in our list. Table 3, therefore, gives 
the numbers of octahedral macles in ascending grades of spread (E/T), 
and Table 4 gives corresponding particulars for the dodecahedral types. 
A comparison of Tables 3 and 4 confirms the evidence of Tables 1 and 2 
to the effect that, excepting at Jagersfontein, octahedral macles have a larger 
spread than dodecahedral ones. Indeed, in the case of two Bultfontein 
specimens included in the numbers of Table 4 the thickness was actually 
greater than the length of edge; and out of sixty-four Bultfontein dodeca- 
hedral macles no less than eight had a spread-ratio less than 1"5. 
Again, of the whole number measured (= 392), considerably more than 
half had a spread-ratio between 2 and 3, whereas only about a quarter had 
a lesser spread. Thus it is proved that the average spread, 2'49, deduced 
