TF. M. Hutchings — Clays, Slates, and Contact- MetamorpJmm. 73 



Staurolite may be appropriately mentioned at this point. It is 

 also not a silicate of alumina only, and from its chemical composition 

 may be reasonably supposed to have an origin of somewhat similar 

 nature to that of cordierite. 



It does not appear to occur as frequently as the other minerals 

 now being considered ; but when it occurs it seems to form in the 

 less intense stages of metamorphism, as for instance in the example 

 studied by Mr. Barrow, to which reference is made below. 



As regards the other minerals, the aluminous silicates, it is usually 

 assumed that they represent the " kaoline " of the slates, etc., and 

 there is on the face of it much in favour of this view. If this be so 

 we should expect I'ocks rich in alkali in proportion to alumina to 

 give little or none of these aluminous silicates during metamorphism, 

 and there are cases where this is borne out, as at Shap for instance. 

 But I do not think that an examination of the analytical evidence in 

 general will show that we can make a rule of this. There are 

 analyses of rocks very rich in andalusite, for instance, which show 

 high percentage of alkali ; and copious develo^xuent of andalusite 

 and felspar may take place together, as is seen in some of the 

 Andlau rocks. At Shap also, we have altered rhyolitic ashes in 

 which there is a great deal of newlj^-formed felspar, together with 

 andalusite, and an unusually large development of sillimanite. 

 Therefore it would appear as if the formation of these minerals is a 

 more complex question than to depend simply on certain ratios 

 among the chemical constituents of the rocks metamorphosed. 



For the rest, looking at their chemical composition and their 

 mode of appearance in our sections, their relationships to other 

 minerals, etc., it would be even more difficult than in other cases 

 to imagine their origination in any other way than by a dissolving- 

 up and complete re-combination of the original materials of the 

 slates, etc. 



Concerning the conditions determining loMcli of the aluminous 

 minerals shall be formed, we have as yet very little to guide us ; 

 but in a paper of exceptional interest, recently read before the 

 Geological Society, Mr. Barrow produces evidence to show that 

 probably difference of temperature is the principal controlling 

 element. In his work on a contact-region in Scotland, Mr. Barrow 

 has observed that sillimanite, cyanite, and staurolite occur in the 

 order named as we recede from the actual contact, and he suggests 

 that the occurrence of each mineral coincides with a zone of 

 temperature. He correlates this observation with the experimental 

 of cordierite in a slide it Tvill nearly always be possible to prove that we have a hiaxal 

 mineral before us, and therefore not quartz. But in a thin section it may often 

 happen that we cannot say whether the biaxal grain is cordierite or felspar, simply 

 by its optic examination. Felspar grains, without any cleavages or definite forms, 

 occur in these rocks. The refraction and bi refraction are not sufficiently different to 

 serve as guides when cut thin, and in many of the best occurrences of cordierite in 

 contact-slates the dichroism of the mineral, and the " halos " round its enclosures, 

 are not discernible at all in most grains. Enclosures, etc., may aid in discrimination, 

 and usually do so ; but I have satisfied myself that the determination is often a 

 difficult one, and think that there is considerable chance of cordierite being over- 

 looked in some cases in consequence. 



