and the lamination of acid lavas. 41 
sealed tubes. The experiments of MM. Friedel and Sarasin,* 
upon the reproduction of orthoclase and albite are especially 
applicable to the case of lithophysze. Among the forms of 
orthoclase produced was that of adular, also observed in the 
hollow spherulites. Along with orthoclase they obtained quartz 
and at higher temperatures tridymite. Other experiments, 
especially those of K. von Crustschoff,f have established the 
same relation between quartz and tridymite when formed in 
closed tubes. Tridymite being produced at a bigher tempera- 
ture than quartz, and frequently both occurring together within 
the same closed tube. 
Magnesia olivine and Peenetite have each been reproduced 
by aqueo-igneous methods, and fayalite is a common product of 
puddting furnaces. So that the group of minerals composing 
the lithophyse in this obsidian are such as may be formed 
through aqueo-igneous processes. 
The well-known experiments of Daubrée,{ on the effect of 
superheated steam on a glass tube, have a special bearing on the 
question under consideration. The glass used by Daubrée 
differed in chemical composition from that of the obsidian, 
but the general nature of the results bears a striking sim- 
ilarity to many characteristic features of lithophyse. The 
anhydrous glass was partly converted into a hydrous silicate, 
accompanied by considerable increase of volume; part was 
reduced to a white mass distinctly fibrous with a delicate band- 
ing parallel to the surface of the glass tube and across the 
direction of the fibres. The surface of the tube was in places 
warped, blistered and excoriated, and often full of cracks. 
There was, also, a delicate foliation parallel to the surface of 
the tube. Under the microscope the altered glass contained 
minute spherulites, microlites, and small crystals of pyroxene 
(diopside), and larger. spherulites, probably chalcedony. The 
surface of the glass was covered with prismatic crystals of 
quartz. The transformation was wholly produced by the in- 
fluence of superheated steam. 
In the case of the obsidian from Obsidian Cliff, chemical 
analyses show a loss upon ignition of 0°66 per cent. That this 
loss is mostly due to water seems probable from the researches 
of MM. Boussingault and Damour, and others, whose very care- 
ful tests have shown the presence of water and a little chlorine 
in a number of obsidians. That this amount of volatile matter 
is sufficient to convert the dense obsidian into pumice like that 
on the surface of this lava flow, is shown by melting a frag- 
* Bull. Soc. Min., 1879, ii, p. 158, and 1880, iii, p. 171; also Comptes Rendus, 
1881, xciii, p. 1374, and 1883, xcvii, p. 290, 294. 
+ Amer. Chem., 1883. Also Tscherra. miner. u. petrogr. Mitth. iv, p. 536. 
¢ Etudes Synthetiques de Geologie Experimentale, 1879, p. 154 to 179. 
