14 SHALER ON THE CLASSIFICATION OP LAVAS 



the conglomerates of the Swiss Miocene, but not infrequently it goes much further, 

 and the pebbles are drawn out into very long forms, often to several times their 

 original length, while the sandy matrix has clearly been found in a semi-fluid con- 

 dition. On the island of Aquidneck in Narragansett Bay, this elongation occurs 

 in particular localities, I believe in all cases near extensive faults, while a mile away 

 the pebbles will be found essentially unchanged in form. Between this pasty condition 

 and the fluidity that would make a dyke stone there is no essential diflference. I have 

 endeavored to show that at certain points the Roxbury conglomerate appears to pass into 

 an amygdaloidal trap, retaining its conglomeratic characters, even where it is distinctly 

 fused, and passing insensibly into a completely amorphous mass.^ Again, at Marblehead 

 Neck, the porphyries seem to pass from an apparently stratified mass into true dyke 

 stones. Instances of this rather indeterminate sort could be multiplied, but as this paper 

 is not intended to furnish a detailed inquiry into the facts that support the propositions it 

 sets forth, I will not go further with the statements. 



While included lavas or dykes represent the less vigorous forms of the forces that in their 

 more intense forms give us true volcanic ejections, they yet differ from them in the 

 evidence of gaseous action. It seems to me reasonable to suppose that there is an incom- 

 plete series of phenomena connecting dykes and volcanos. When the ejection serves only 

 to fill the vacuum of a fault or gash fissure, and has not the impulse necessary to force aside 

 the obstructions and make its way to the surface, we have an ordinary dyke ; when the 

 uprush of expanding gases is great enough to break a way to the air, the result is a vol- 

 cano. 



The question arises, is it likely that the ordinary dykes which are so abundant in our 

 Archaean rocks, have in many cases, been the tubes up which came the products of volcanic 

 eruptions ? A little consideration makes this appear improbable. In the first place even the 

 smaller class of volcanic cones have vents some hundreds of feet in diameter, while it is 

 unusual to have a dyke of such size. Moreover, we must believe that the long continued 

 passage upward of the volcanic products through a fissure, would necessarily bring about 

 a great change in the character of the walls that bound it. They would be to a great 

 extent melted, and could not help showing marks of the strong uprush of the volcanic 

 products. In fact our dykes almost always have their walls so little disturbed, that we can 

 trace the corresponding sides of the break for great distances, and the change in the coun- 

 try rock from heat, is usually singularly small. This shows pretty clearly that our dykes 

 are not often to be classed as volcanic channels. 



It may be noticed that we sometimes find evidence that the paths taken by dykes after- 

 wards become more opened, and veins take their place beside the dyke, showing 

 that even when the temperatures are high enough to enable the penetrating waters to 

 carry gold, silver, and other metals, the opening of a fissure may not give passage to 

 volcanic matter. This would seem, at first sight, to militate against the view of the 

 ori<nn of dykes presented in the preceding propositions, for the temperature, if sufficient 

 to enable water to convey such metals, would seem to come very close to that necessary to 

 melt the most fusible substances. But we do not know the actual temperatures at which 

 water can carry the various metals. They may be soluble in temperatures much below 



1 See Proc. Boston Soc. of Nat. Hist., xx, 129. 



