490 proceedings: geological society 



gestively similar to species found in the Caloosahatchee marl of 

 Florida. 



R. T. Hill spoke of work which he had done years before in the 

 West Indies, and expressed gratification that the New York Academy 

 of Sciences had taken up the Porto Rican field. He mentioned some 

 of the structures which had been exhibited in the views and advanced 

 the opinion that the islands of Porto Rico, Santo Domingo, and Cuba, 

 with the connecting banks, represented horsts left at a high relative 

 elevation by the down-sinking of faulted blocks at the sides. Sidney 

 Paige inquired about the structures shown in the San Juan formation, 

 whose origin Fenner had ascribed to the consolidation of old sand- 

 dunes. He thought that for such an origin the cross-bedding should 

 be of a somewhat different type than that shown. Fenner replied 

 that the reference to such an origin had been made because of the areal 

 distribution of these hills along the north coast and because of the 

 internal structures, which Avere difficult to account for otherwise. 

 A. C. Spencer inquired as to whether any mineral deposits had been 

 seen and as to the mineral resources of the island. Fenner replied 

 that indications of mineral were known at various places and several 

 of these had been visited. The only active work being done was a 

 little gold-washing by natives along streams, but prospects had been 

 opened up in gold, copper, lead, and iron. 



R. B. Sosman: Types of columnar structure in igneous rocks (il- 

 lustrated). From the physical standpoint two principal types of 

 columnar structure may be distinguished. The first and most common 

 is that due to contraction of the crystallized rock during cooling, where- 

 by strains and stresses are produced which yield different sizes and 

 kinds of prisms according to the magnitude of the temperature gra- 

 dient, the rate of cooling, and other factors. The second type arises 

 from convectional circulation of the still liquid rock. Experiments 

 on low-melting materials have demonstrated that the vertical transfer 

 of heat by convection is capable of dividing a liquid into hexagonal 

 cells which leave their record in the crystallized mass and control its 

 further division by contraction into hexagonal columns. The two 

 methods produce prisms which differ in attitude, in the relation of di- 

 ameter to length, in the frequency of 4, 5, 6 and 7-sided polygons, in the 

 frequency of certain angles, and in the type of their cross-jointing. 

 Attention was especially called to the usefulness of quantitative data on 

 prismatic structures in igneous rocks as an index to the original con- 

 ditions of occurrence of the rocks containing these structures. 



C. N. Fenner, 

 C. H. Wegemann, Secretaries. 



The 297th meeting was held in the lecture room of the Cosmos Club 

 on May 12, 1915. 



informal communications 



R. B. Sosman: Two suhordinaie types of prismatic structure. In 

 addition to the two principal types discussed at the meeting of 

 April 28, two subordinate types should be distinguished. The first 



