230 Reports and Proceedings — Tlte Royal Society. 



underground cliannels and springs, are very fully described, with 

 remarks on sources of contamination. 



4. Iridosmine in the Transvaal. — This alloy was discovered 

 several years ago by Mr. A. F. Crosse in the Rand bankets. Its 

 occurrence in the New Kietfontein Mines is recorded by Mr. C. Baring 

 Horwood (Trans. Geol. Soc. South Africa, xv, 1912), and he describes 

 it as an intimate mixture of iridium and osmium, generally with 

 ruthenium, and sometimes with a little platinum, chromium, gold, 

 and palladium. The metals in the banket reefs, in his opinion, are 

 of secondaiy origin, having oi'iginated as segregations formed by 

 magmatic concentration in the basic eruptive rocks of the mines, and 

 from these dykes they were extracted by the action of superheated 

 gases. Later still, hydrothermal action probably played an important 

 part in concentrating them in the banket reefs. 



I. — The Royal Society. 

 March 6, 1913. — Sir Alfred Kempe, Vice-President and Treasurer, 



in the Chair. 



The following communications were read : — 



1. "The Evolution of the Cretaceous Asteroidea." By W. K. 

 Spencer. Communicated by Dr. A. S. Woodward, F.R.S. 



An endeavour is made to ti'ace the evolution of tlie Starfish through 

 the whole of the Cretaceous deposits. At the first sight the material 

 appeared to be unpromising, for complete or even fragmentary 

 specimens are rare. It has been found possible, however, to use 

 the isolated marginal plates which are found fairly commonly on 

 weathered chalk surfaces. It is shown that these marginal plates 

 have a shape and ornament characteristic of each distinct species. 

 Tlie species may be arranged in lineages, and the examination of 

 large collections made l:)y English and Continental workers make it 

 feasible to trace the life-history of most of the lineages. Each 

 lineage shows definite stages of advance (elaboration) followed by 

 stages of regression. Occasionally a lineage will re-elaborate after 

 passing through the regression stages ('rhythmic' or 'periodic' 

 variation). Elaboration and regi'ession are closely parallel in the 

 lineages. Thej^ consist of changes in the size, the height, and the 

 character of the ornament of the plates. 



Variation as shown, both through measurements and by general 

 examination, is, in general, of the 'continuous' type. Discontinuous 

 variations (' saltations') are known, but they appear to be relatively 

 of little importance. The measurements also show (1) that there is 

 close connexion between the general vitality of the race, as shown by 

 its range of variability, and the rhythmic phases alluded to above ; 

 (2) that the rate of elaboration is affected by environment. 



It has not been found possible to show that selection plays any 

 predominant part in the evolution of the lineages, and it is suggested 

 that inborn racial character is the predominant influence in causing 

 modification. 



