86 



Prof. H. E. Armstrong. 



[Mar. 5 3 



in the adult bird the cones are most numerous. Some of the 

 bipolar cells from the internal nuclear layer emigrate through the 

 external molecular layer, and develop into young cones. These 

 investigations on the retina led to an inquiry as to what relationship 

 there might be between the rod and cone cells, and the ciliated 

 epithelium of the cerebral ventricles, and it is found that these are 

 homologous structures. The germinal cells which line the cavities of 

 the cerebral vesicles remain there after they cease to divide, and from 

 them cilia develop. The inner ends of the myelospongiirm fibres do not 

 develop cilia as is usually described. 



The processes of the hexagonal pigment cells appear on the same 

 date and at the same spot as the rods and cones and do not at first 

 contain pigment granules. They resemble pseudopodia in their mode of 

 development, and also, after full development, they behave like pseudo- 

 podia, for they elongate and retract under the influence of light and 

 carry the pigment granules with them (the granules having no inherent 

 power of movement). 



The inner wall of the retinal cup has been seen to have the same 

 structure as the wall of the cerebral vesicle in the early stages, and 

 even later (as at the 8th day) some resemblance can still be drawn 

 between them. 



" The Classification of the Elements." By Hexey E. Armstrong, 

 Y.P.E.S. Eeceived March 5 — Bead March 20, 1902. 



Although no direct evidence acceptable to chemists has been adduced 

 which in any way justifies the belief that the elements are decomposible, 

 it is impossible to resist the conclusion that they are genetically related 

 — so closely in many respects do they resemble a series of related com- 

 pounds, especially when regarded from the point of view of the organic 

 chemist. The generalisation known as the Periodic Lav: is in itself a 

 justification of this view : the manner in which interrelationship 

 becomes manifest when they are classified in accordance with its canons, 

 being probably the strongest of all the arguments which can be cited 

 as tending to show that the elements are compounds — but compounds 

 very different from those with which we are accustomed to deal. Even 

 in the form in which it was put forward by MendeleefF, however, the 

 periodic generalisation is but a first approximation : and the great 

 Eussian has himself pointed out that it needs improvement and develop- 

 ment.* As chemists are beginning to recognise this,f I venture to 

 submit a scheme of classification which I have been led to draw up in 



* Faraday Lecture, ■ Trans. Chein. Soc.,' 18S9. p. 656. 



f Comp. Biltz, ' Deut. chem. Ges. Ber., 5 1902, p. 562. 



