RHYTHMIC BANDING OF MANGANESE DIOXIDE 615 
of some manganese mineral. No definite evidence of the presence 
of such a mineral could be found, although one would be supposed 
to be occupying, or to have occupied, the center of the structures. 
The original mineral may have been spessartite, which, as has been 
noted, occurs in rhyolites, as at Rosita Hills, Colorado. It might 
also have been some manganese-rich ferromagnesian mineral, such 
as biotite, or one of the pyroxenes or amphiboles. 
Water could readily enter the porous tuff and attack the spes- 
sartite. The manganese would be taken into solution largely by 
the water from the immediate rainfall and probably in the form of 
the carbonate. It could be carried out some distance (about one 
inch in the larger structures) before deposition ceased. Deposition 
would be caused by the oxidation of the manganese salt to manga- 
nese dioxide, a process that would be favored by the arid climate 
of the region. . 
A study of the bands favors the view that they are due to the 
same process as is the formation of Liesegang’s rings, also known 
as rhythmic banding.t Once the manganese is in solution it will 
diffuse outward through the tuff. Holmes’s demonstration that 
the rings may be produced in loosely packed flowers of sulphur 
shows that gels are not essential for their development. The porous 
tuff would be analogous to the flowers of sulphur in permitting the 
diffusion in this case. 
The manganese solution would diffuse outward at a rate de- 
pendent upon its concentration. As it mingled with other solutions 
containing oxygen, the concentration of the respective ions would 
increase until a labile condition occurred, when precipitation would 
take place, thus producing aring. ‘The color of this ring would de- 
pend upon the concentration of the solution and in part upon the 
hydration of the resulting oxide. The manganese solution would 
then move on (provided its rate of diffusion exceeded that of the 
other solution) through the zone depleted of oxygen until a second 
t Recent discussions of rhythmic banding or precipitation are given in the follow- 
ing papers: J. Stansfield, “‘Retarded Diffusion and Rhythmic Precipitation,’ Am. 
Jour. Sci., 4th series, XLIII (1917), 1-26; H. N. Holmes, “Rhythmic Banding,” 
Science, New Series, XLVI (1917), 442; J. Stansfield, ‘Rhythmic Precipitation,” 
ibid., New Series, XLVII (1918), 70. 
