Editorial Comment. 323 
island, accompanied by a great noise like the cannonading of heavy 
guns; and the next day the flames and the earthquake continued. The 
flames and smoke were seen for a long time. Many masses of pumice 
stone were ejected on the first appearance of the island." 
Professor Davidson adds : "At that time it was, perhaps, 
only one-quarter the size of its present dimensions ; and it in- 
creased in size, growing higher, and breaking down at the same 
time on all sides. Finally, about 1823, it seemed to have be- 
come unchangeable. Until it ceased to increase in size it was 
hot, as well as the sea water around it ; while smoke and steam 
rose from it continuously." The lower figure in plate xxi 
shows the new island as represented by Davidson. It was larg- 
er than the original island. It was more or less enveloped in 
steam and a "black smoke issuing from it as if tar were burn- 
ing." * * * "The sides are very steep ; and, apparently it has aris- 
en from the depths without developing outlying dangers, be- 
cause with a heavy swell running, no breakers were seen. It 
rises very steeply with a rough ogee curve." This island is 
about half a mile north-northwest of the old one, and about 
1000 feet high. 
According to W. H. Dall there are other Aleutian islands 
of exactly similar origin, notably Koningi and Kastochi, and 
Pinnacle island near St. Matthew island. Of some of these he 
gives profile sketches. Science, Jan. 25, 1884, p. 92). 
In the western part of the United States are numerous 
peaks of volcanic rock that have been described as volcanic 
necks and plugs. How many of these can be certainly includ- 
ed under the term pelelith it is impossible to state. The new 
fact of the formation of a volcanic plug by the slow ascent of a 
mass of volcanic rock already solid, from below the crater 
rim, as at Martinique, renders it necessary to hesitate in ascrib- 
ing any manner of origin to such volcanic necks as are well 
known, and to institute fresh examinations with view to com- 
parison with that of Mont Pele. It has been customary, follow- 
ing Dutton, to infer that volcanic peaks of this nature have 
simply been uncovered by circumdenudation, involving in some 
cases the removal from the vicinity and from the surrounding 
region, embracing sometimes thousands of square miles, a 
thickness of three or four thousand feet of the sedimentary 
strata. Not to mention the difficulties that arise in some places 
in the efifort to apply such a theory, it may be asked whether 
