268 Russell — Massive-Solid Volcanic Eruptions. 



other hand, correspond fully as well with the similar features 

 of many volcanic necks. The chemical composition of the 

 rock of which the Peak is composed, so far as suggestive of its 

 degree of fusibility, is similar to that of the material erupted by 

 Pele ; and so far as this fact has weight, it is evidently not 

 opposed to the idea that the rock in question was extruded in 

 a massive-solid condition. The well-defined columnar struc- 

 ture of the Peak, however, is so similar to that of many vol- 

 canic necks and still more numerous igneous dikes, and is so 

 unlike the jointing observed in at least two examples of mas- 

 sive-solid extrusions, and so unlike, also, as we seem justified in 

 assuming, the jointing to be expected in all such extrusions, 

 that it favors the conclusion of its being a volcanic neck or a 

 portion of an igneous dike, rather than of the nature of the obel- 

 isk of Pele'. 



In reference to the criteria by means of which a distinction 

 can be drawn between residual masses of lava sheets left as 

 columns on the tops of hills or buttes, and obelisks of the Pele 

 type, but little need be said, since even a cursory examination 

 of such monuments of erosion is usually sufficient to reveal 

 their history. Erosion columns may be composed of any vari- 

 ety of volcanic rock, but as is well known, are most commonly 

 basaltic, are usually vertically jointed and frequently conspic- 

 uously columnar, and rest on soft or incoherent material. In 

 most instances, also, in regions where one such residual column 

 occurs, others of similar nature are apt to be present, as well as 

 flat-topped mesas and even broad table-lands. Thus in the 

 nature of the material of which the columns referred to are 

 composed, and in their structure and associations, they differ 

 widely from massive-solid extrusions. 



It is at present impracticable to apply the above mentioned 

 criteria to the columns in the Deccan trap regions of India, 

 referred to by Strachey, except so far as topography and the 

 general nature of the rock are concerned. Topographically 

 the columns represented in the sketch mentioned on an early 

 page of this essay, seem to agree much more nearly with ero- 

 sion columns than with volcanic obelisks, a view which is sus- 

 tained by the presence of two examples near each other. The 

 Deccan trap is composed of basaltic rock, such as is common 

 in erosion columns and not as yet known to occur in volcanic 

 obelisks. The evidence seems, therefore, to indicate that the 

 columns in question are residual masses left by the nearly com- 

 plete erosion of a formerly widely extended sheet of lava. 



From the considerations presented in this essay, it will be 

 seen that the recent eruptions on Martinique have made impor- 

 tant contributions to both geology and geography. In the 

 investigation of volcanoes geological and topographical studies 

 go hand in hand and mutually assist each other. 



