388 Tlie AlllCl'icaii (r\(>lnt/t.st. HtciinlMT, 1891 
tli:it llic tiltiTs :iii«l sc;il('s ill tlic uroiiiid iiniss of |K>r|ili\ rii's iind 
rlivolitcs niiiv t>c iiiimitc itiirliclcs (tC fcl»ls|):ii- iiidctcriiiiinililc hv 
oxistinji" iiu'iiiis of rcsciircli. 
In ;\ similar inniincr the tt'iiii •pctiosilcx is ciiiploycd liv the 
IVrciicii |)('ti(>>ir:iplu'is. iiiidcr the U':i<l of .Michel-Lev y. in a loose 
way. as is adniirahly illustrated liy the (U'tinition. '-a partially 
amorphous mairma impregnated with silica already individualized 
in a state of opal or chalcedony. " Spherulites aic supposed. l»y 
the same authority, not to consist of known minerals, unless they 
be ((uartz and feldspar, interiri'own in the manner of peoniatite. 
'I'lie rt'sult of tlu' author's study of the Colorado material seems 
to show that indetinite sulistances. such as crystallites. [>etrosilex 
and microfelsite. have been assumed t() l)e present unnecessarily 
in many cases heretofore, iiiidei' the iiitliieiice of preconceived 
ideas. 
The clu-mical constitution of the i)itclistoiH' containinji' the Col- 
orado splierulites shows, that on complete crystallization the rock 
resultiuii' would l)e comi)osed nearly two-thirds of alkali feldsj)ar 
and a little more than one-third of free silica. 'i'he author coin- 
ci<les with ld<liniis in the description of Obsidian cliff, that, in- 
stead of l)eino; due to an arrest of crystallizinu' consolidation (sup- 
posed ))V :dl European authors) spherulites of the smallest size as 
well as the larucr spherulitic masses, are due to the crystallization 
of some definite minerals from a nuiiiina. under special conditions. 
There are a few older crystals in these rhyolites which, at a })eriod 
prior to the s[)heiulitic. wert' formed from the magma and by 
their arranu'cment liring out the Huidal structure. These took up 
the small ix-rci-ntaues of lime, magnesia and iron oxides which 
analysis shows to I'xist in the pitchstone, forming phenocrvsts of 
plagioclase. h'avi's of biotite. microlites of augite and trichites of 
magnetite. ^Ir. Cross examines specially the larger spherulitic 
masses, but he considers even the sinalicst to consist of di'finite 
and often identifiable minerals. 
As to the origin of these spherulites. Mi-. C'ross supposes that 
amorphous silica plays an important role. I"'irst a colloidal sub- 
stance, embracing the elements of silica and of feldspar, is sup- 
l)osed to l»e separated from the magma, this being '-a local change 
in the character of the magma.' but whose cause and attendant 
conditions Mr. Cross does not attempt to state. From this pri- 
mary colloidal globule are geneiated amorphous silica in one di- 
