136 CARNEGIE INSTITUTION OF WASHINGTON. 



cedure is given which shows that it is, if anything, more accurate than when 

 the thermal head is observed. It evidently releases the experimenter's 

 time for more essential observations. 



These more numerous observations contribute to precision when the 

 heating current is relatively unsteady. In extreme cases, however, a coulom- 

 eter is almost unavoidable, probably with some loss of precision compared 

 to the usual method at its best. 



The heat conductivity of the leads to the electric heater, and the heat 

 produced in them, may give rise to errors, for whose magnitude approximate 

 formulas are given. The most efficient dimensions for the heater leads 

 vary greatly with the other conditions of the experiment. 



(480) The origin and mode of emplacement of the great tuff deposit in the Valley of Ten 



Thousand Smokes. Clarence N. Fenner. National Geographic Soc, 

 Contributed Technical Papers, Katmai Series, No. 1. 1923. pp. 1-74. 



During the period of volcanic activity in the Katmai region in 1912 the floor 

 of the Valley of Ten Thousand Smokes became covered with a deposit of tuff 

 of great thickness, possibly attaining several hundred feet in places. Certain 

 features of this deposit have been described in previous publications, but 

 not in detailed or complete form. As a result of the Katmai Expedition of 

 1919, certain very definite conclusions regarding its origin were arrived at, 

 and these conclusions and the reasons for them are given in detail in this 

 article. The principal points that seem to be established by the evidence 

 are that Katmai crater could not have been the source of the material, nor 

 could any other single point of extrusion have supplied it all, but that a 

 number of vents (possibly a large number) must have broken out on the floor 

 of the valley, in Katmai Pass, and perhaps on the lower slopes of Mount 

 Trident. Probably the new volcano Novarupta was a principal vent, but 

 others are indicated along the fissures which shatter the floor of the valley 

 and along which the present fumaroles are situated. The material was at a 

 high temperature when extruded. In many respects the form of eruption 

 doubtless resembled very closely the nuees ardentes of the Antillean erup- 

 tions of 1902. Certain characteristics of the fissures of the valley area 

 lead to the inference that a sill-like body of magma was driven under the 

 valley floor. 



The make-up and general character of the deposit are described. Most 

 of the material is a highly siliceous glass, shattered to a fine dust, or having 

 the form of pumice, but mingled with this is a considerable amount of basic 

 scoria and pumice. To account for the latter, we are led to a study of the 

 material that was explosively ejected from Novarupta after the sand-flow 

 occurred, and of that which forms the dome. This study shows that a large 

 amount of old basic lava became involved in the newly rising siliceous magma 

 at Novarupta and was melted down by it. Similar processes are supposed 

 to have occurred at other vents, and by this melting down and by subsequent 

 inflation the basic pumice and scoria were developed. 



A comparison is made with other instances of nuees ardentes, and the 

 characteristics of this form of eruption are discussed. 



(481) The crystal structure of the alums. Ralph W. G. Wyckoff. Am. J. Sci., 5, 209-217. 



1923. 



This paper is a briefer account, in English, of the experiments described 

 in No. 481a. 



(481a) Die Kristallstruktur der Alaune. Ralph W. G. Wyckoff. Z. ICryst., 57, 595-609. 

 1923. 



It is shown that Laue photographic and spectrum photographic studies 

 give T h*^ rather than the previously assigned T h^ as the corresponding space 



