426 MR. F. EUTLEY OX COMPOSITE SPHERULITES 



tliat between these small spherical pellets of tridymite vacuities 

 still remained. In such empty spaces he finds that in most 

 cases emery or other grinding-material has collected during the 

 preparation of sections. In the central portion of one of the frac- 

 tured spherulites in the specimen now described, he detected one or 

 two small crystals of altered fayalite. 



Although it is not very easy to give the results of a conversation 

 accurately, without having taken notes at the time, yet I thiuk that I 

 have here expressed the opinion with which Mr. Iddings favoured 

 me, clearly and correctly, and he kindly gave me permission to 

 make this statement. 



That it is difficult, if not dangerous, to question the ojnnion of a 

 caretul observer who has had such unparalleled opportunities for the 

 study of spherulitic structures as Mr. Iddings has enjoyed, is a 

 fact of which I am fully cognizant ; yet there appear in this particular 

 instance to be certain matters of detail which seem to support the 

 conclusions expressed by me in the foregoing paper. 



They may be summarized in the following manner : — 



1st. That although diminution in \olume is attendant upon crys- 

 tallization in cases such as that with which we are now dealing, 

 w^ould it occur to an extent so appreciable that a cavernous or 

 spongy structure would be thus developed tliroughout the entire 

 mass of a spherical body an inch in diameter, or would it not rather 

 result in the production of one or two appreciably large fissures or 

 cavities such as are met with in lithophvsae, assuming that those 

 fissures or cavities would not be developed through other causes 

 than contraction ? 



2nd. That if such a spongy condition of the spherulitic body were 

 induced, so that the entire spherulite consisted merely of divergent 

 and anastomosing filaments composed of felspar microliths, forming 

 a kind of delicate network, the interspaces of wliich were filled 

 neither with vitreous matter nor with any other solid, would 

 tridymite deposited within such interspaces be developed in the 

 form of closely-packed, well-defined, spherical pellets, or would it 

 t(md to crystallize out along the twigs of the divergent microlithic 

 structure of the spherulite, thus filling all the interspaces com- 

 pletely ? 



3rd. That tridymite is in the habit of forming more or less 

 spherical groups is well known, but would such spherical groups be 

 formed in a closely matted aggregate of crystalline rods, since each 

 small pellet (primitive spherulite) would consist as much of felspar as 

 of tridymite? Purthermore, would not these minute felspathic rods 

 afi'ord surfaces along w^hich crystallization would be set u]), and thus 

 seriously impede the crystallization of the tridymite in spherical 

 aggregates, if, indeed, it would not render such spherical development 

 impossible? 



4th. The small si)herical pellets have well-defined boundaries 

 such as ordinary spherulites might have. Examination in polarized 

 light reveals the presence of comparatively little, if any, isotropic 

 matter, the pellets appearing to consist to a very laige extent 



