284 
GILBERT. 
crevice gave way, letting what lay above it pour into the 
abyss below. Possibly there was moisture in the crevice, 
and the inrush brought heat enough to cause explosions of 
steam. If the floor fell in, the pits should be rimless; if it 
was blown out, they should be rimmed. I was not able to 
satisfy myself as to their actual character, and recorded 
observations are discrepant, but the features are not too 
minute for accurate determination. 
In a part of the rill of Hyginus the pits are set so close 
that their edges adjoin; other rills are composed wholly of 
pits, and these lead by gradation of characters to rows of 
separate pits where no rills are visible. Fine illustrations 
of the last may be seen along the western base of Copernicus, 
almost half way to Eratosthenes. If my conjecture is cor¬ 
rect, these mark the line of a fissure that was filled by a 
molten flow connected with the formation of Copernicus. 
The rills that have no visible bottoms, but are seen only 
as black lines, are the unfilled fissures necessary to complete 
this series of features at the opposite end. 
White Streaks .—The only remaining great group of features 
are the white streaks. These are bands of color, sometimes 
faint, sometimes brilliant, but always indefinitely outlined, 
like the tail of a comet, and some of them stretch for long 
distances across the moon’s surface. Their courses are inde¬ 
pendent of the configuration. They pass up and down the 
slopes of craters without either modifying their forms or 
being interrupted by them. The more prominent of them, 
and probably all, occur in systems, and those of each system 
radiate from some crater. This crater is itself lined with 
white and is usually more resplendent than the radiating 
streaks. We need not take time to consider the various con¬ 
jectures which have been published concerning their origin. 
It suffices to say that all but the least plausible ot these con¬ 
jectures have been advanced as suggestions merely and have 
not been fully endorsed by their authors; but there is an 
unpublished suggestion, made by Mr. William Wiirdemann, 
of Washington, which is at once so apt and so simple that I 
