196 Proceedings of the Royal Physical Society. 



build up successive shells of a hemispherical form ; and as 

 the composition of each layer often differed, within certain 

 limits, both in structure and composition, from those with 

 which it was associated, these shells, when subsequently cut 

 and polished, usually exhibit a series of concentric rings of 

 different colours, which may be likened to eyes. In this 

 manner are Eyed Agates formed. There are few agates 

 which do not show at least some traces of these. 



Usually these " eyes " are formed at the stage just referred 

 to, but the process may occasionally be repeated at various 

 stages later in the developmental history of the agate, with 

 much the same result. 



Where '* eyes " occur in an agate, it will be seen that after 

 the hemispherical stage has been reached, the succeeding 

 layers, if of Chalcedony, tend to conform more and more to 

 the general shape of the interior of the cavity, so that the 

 successive zones over each eye gradually change their form 

 as they approach the centre, passing first into layers that are 

 nearly flat, and then becoming slightly concave towards the 

 centre as they grow beyond the influence of the earlier 

 irregularities. 



Where the growth of two or more eyes brought them 

 into contiguity, the form of the layers deposited between 

 them tended first to assume that of sharp re-entering angles, 

 and then, with further growth towards the centre, caused 

 them to merge into simpler curves. Cross-sections of an 

 agate so formed are known as Fortification Agates, from 

 the fancied resemblance of these salient and retiring angles 

 to the plan of certain forts. 



In all the foregoing notes on agates it has been assumed 

 that the solutions of silica carried in by osmosis have been 

 deposited very slowly, and in very thin layers at a time. 

 Many agates which happen to have been formed of layers 

 varying in composition and in colour show this very well. 

 It is not uncommon to meet with examples in which several 

 hundred layers can be counted in the thickness of a single 

 inch, and even these layers themselves may consist of 

 numerous others of greater tenuity still. 



It has further been assumed for convenience of descrip- 



