FEBRUARY 1, 1884.] 
compact mass. On the 11th and 12th came freezing 
weather, and the fall of a small amount of very light 
snow. On the 13th the thermometer, toward noon, 
rose above freezing-point, with a stiff breeze from the 
south. This wind so acted on the surface-particles 
of the upper layer of uncompacted snow as to set 
some of them in motion. Each particle thus set in 
motion, owing to the soft condition of the surface- 
snow, formed a nucleus, which, as it proceeded, 
forced along by the wind, gathered the contiguous 
portion of the soft layer, and assumed the form of a 
cylinder, with a conical cavity at each end, and hav- 
ing a length about twice as great as its diameter. 
The size depended upon the inclination and smooth- 
ness of the surface traversed. The largest cylinders 
I saw were about three feet long, at which limit they 
acquired sufficient weight to indent the frozen sur- 
face of the under or main body of the snow. This, 
of course, stopped the further rolling of the mass. 
The only locality where they acquired the above large 
size was where the surface had a slight inclination to 
the north; and the snow was deep enough to cover all 
weeds, leaving a perfectly uniform and smooth sur- 
face for their formation. In many cases the balls 
were rolled up an inclination of as much as one foot 
in ten, when exposed to the unbroken force of the 
wind; but those thus formed acquired weight suf- 
ficient to resist the pressure of the wind, when about 
six inches in diameter. When the surface inclined 
toward or directly away from the wind, the balls 
traversed a straight path; but, when the surface de- 
clined to the north-east or north-west, the path was 
a curve; at its initial, approximately straight; but, 
as the ball acquired weight, its direction was a com- 
promise between that required by gravity and that 
by the direction of the wind, until, in some cases, the 
ball obeyed gravity alone. The most curious part of 
the display was the abundance of the balls. While 
travelling three miles, I saw what I estimated at over 
a hundred acres dotted more or less thickly with 
the cylinders. In some cases there were twenty-five 
balls to the square rod; in others, only two or three; 
averaging, perhaps, eight or ten. I saw multitudes 
in the process of formation, which was as sudden as 
a flash; but they almost immediately assumed a slow 
rate of motion, about that of a mole taking his leis- 
urely walk. Ina few cases the cylinders would stop, 
and afterward be forced into motion again. The 
largest examples required for their formation the 
traversing of from two to three rods. SAM HUSTON. 
Richmond, O., Jan. 16, 1884. 
The wind performed a very pretty feat in some por- 
tions of northern Ohio on the morning of Jan. 13. 
Loose bits of snow were caught up as a nucleus, and 
rolled along upon the surface until balls of consider- 
able size and peculiar shape were formed. The whole 
surface was strewn with the balls; but they were most 
abundant upon lawns and fields where the wind was 
not obstructed, every square yard, in some places, bear- 
ing a ball of greater or less size. The largest observed 
here were upon the college ball-grounds, where they 
reached ten inches in height, and a horizontal length 
of eighteen inches. Even these were swaying as the 
gusts passed over them; and their tapering track could 
be plainly traced back towards the south-west, twenty- 
five or thirty feet, to the apex where they started. 
Their shape was cylindrical, deeply hollowed at both 
ends, so that they looked like ‘muffs,’ and the spiral 
formed by the successive layers was finely regular and 
distinct. 
The meteorological conditions which made the phe- 
SCIENCE. 
115 
nomenon possible were as follows. Two days before 
the occurrence a slight crust was formed upon the 
snow. On the following day an inch of light flaky 
snow fell upon this crust. Then followed the warm 
south-west wind on the morning of the 13th, which 
brought the upper layer of snow into the adhesive 
state, and rolled the balls before the crust was weak- 
ened; the crust sustaining the balls, and keeping them 
up to the wind, and at the same time furnishing a 
smooth floor upon which they could be propelled. 
The nuclei of the balls were obtained from chance 
foot-tracks, walk-borders, lumps blown from trees, 
etc., though often it was difficult to account for them. 
The balls were most abundant and perfect at about 
nine o’clock A.M. Before noon the crust had been at- 
tacked, and all sunk to rounded, insignificant clumps. 
Oberlin, O. ALBERT A. WRIGHT. 
[Similar snow-rolls were seen at Sharpsville, Mer- 
cer county, Penn., on the same day, by J. M. Good- 
win. | 
Halos round the moon. 
On the evening of Jan. 12, at 9.20 (90th meridian 
time), my attention was called to a peculiar appear- 
ance about the moon. The sky was quite clear at 
the time, and there appeared around the moon sev- 
eral colored circular bands. The first was of a bright 
silver-gray shade, and about two diameters of the 
moon in width. The next was yellow, the next faint 
orange, and the next violet. The three bands were 
each about one-half a diameter in width. The outer- 
most band was of a green shade, and about two diam- 
eters in width. At ten o’clock the innermost light 
band remained, but all the others had been replaced 
by a blue band lighter than the surrounding sky. 
H. A. HUSTON. 
Lafayette, Ind., Jan. 14, 1884. 
Explorations in Guatemala. 
Looking over the back numbers of your esteemed 
journal, I came across a slight error. In the article 
‘Lorillard City’ it is said (ii. 412), ‘‘M. Charnay 
found the ruins of an ancient city, which he named 
after his generous patron. In his exploration here, 
he was assisted by a young Englishman, Mr. Alfred 
Maudslay, with whom he shares the honor of discov- 
ery,’’ etc. 
Neither Mr. Maudslay, who arrived at these ruins 
before Mr. Charnay, nor the latter, can claim this 
honor. In fact, Mr. Maudslay distinctly states (p. 
196 of the Proc. roy. geogr. soc., April, 1883) that they 
have been discovered by Mr. Edwin Rockstroh, tutor 
on the Lyceo nacional at Guatemala City. This gen- 
tleman made, during the first half of 1881, a geograph- 
ical and archeological exploration in the northern 
and western parts of the republic, visiting Tikal, and 
navigating the Rio de la Pasion, Kio de Jas Salinas, 
Rio de los Gacandones, and the Usumasinta as far 
down as the ruins mentioned. He sent a short ac- 
count of this voyage to Petermann’s Mittheilungen 
(1881, p. 396). 
In that account Mr. Rockstroh mentioned particu- 
larly the building described by Mr. Maudslay on p. 198 
of the geographical society’s proceedings; and (1882, 
on p. 435) he clearly states that Charnay’s ‘ Lorillard 
City’ is the same as that discovered by him in 1881. 
Mr. Rockstroh mentioned in his first letter to the 
Mittheilungen (July 19, 1881), that the Gacandones 
call these ruins ‘Menche,’ and promised in his last 
notice (1882, p. 435) an explanation of thisname. I 
am not aware that he has furnished one. 
I find in the ‘Historia de la provincia de San 
Vicente de Chiapa y Guatemala,’ by Antonio de 
