lullforliiJ ( 'oiiitNi'iif. ^5in 
anust atiaiii 1k' luhiiitted that the very act of vaporization of water 
when })roiight into contact witli a molten magma, would al)stract 
licat locally from the magma, and that the local differences in the 
heat of the magma \v(^uld extend pari passn with the production 
and ditt'usion of the vapor of water. Indeed these two progressive 
changes would almost exactly complement each other, and hence. 
;i.gain, the presence and the distribution of hydric vapor ••through 
■considerable masses of the magma" would not perhaps be any 
;more une([ual than the distribution of heat. 
It must be understood, however, that here we are considering 
•••considerable masses of the magma,"' throughout which compared 
with other con.si<lerable masses there may or may not have been 
differences of heat and hydric vapor. This is ([uite a different 
thing from tlie assunii)tion of centres, or points (for each spherulite 
is sujjposed to have l)egun its growth at a point), disposed irregu- 
larly through the molten magma, at which there was, temporarily oi- 
<'<mtinuously a greater amount of watery vapor than throughout the 
remainder of the mass. Tlu'se spherulites are immensely numer- 
ous. Some are large, anil sonn^ are microscopically minute, and 
sometimes they t()uch each other. There must, under the hypoth 
esis <»f ^Ir. Iddings. have been an immense num])er of individual 
points, almost an infinite number, at which, though separated 
from each othei- by inierosopic distances, there were maintained 
greater amounts of hydric vtipor than in the intervening spaces. 
Such a conception of the manner of distribution of hydric va- 
por is Mni([ue and c:ui hardly Ite accepted under the well-known 
laws of the diffusion of gases. 
We irannot see as <'ithi;r of these authors has suggested — we 
will not say the cause but — the' essential condition precedent of 
the fornuition of the spherulites in acid lava. Mr. Tddings" re- 
(luirement that within the area of the developing spherulite there 
must have been greater molecular mobility than in the surrounding 
magma, is a step toward the solution of the prol)lem. Whether 
this greater mobility dei)ended on a colloidal secretion from the 
nuigma.envelopingthe point at which the crystallization began. oron 
a point or a globnle. within the nuigma. which enjoyed a preponder- 
ance of hydric vapor, or of greater initial heat, each one of Avhich 
«eems to us untenable, it is phiin that the authors have made valua- 
bleaiul ci'cditalile conti-iliutions to .\nieric:in pet rology. and tli:it in 
