188 H. Gibbons on Springs and Streams in California. 
summit and increased considerably in its descent. The dry 
season commenced rather earlier than common, not enough rain 
to lay the dust falling after the first week in April. My little 
rivulet continued to murmur refreshingly by the roadside until 
July, when it disappeared in places; and by the beginning of 
August it formed a chain of swampy spots and pools. At the 
end of August, when I expected to find it almost desiccated, 
judge of my surprise on encountering a brisk streamlet about 4 
mile from the top of the hill, at a spot which had been perfectly 
dry. From that time it steadily increased, the pools being con- 
nected for the greater part by a continuous stream, by the 10th 
of October, though no rain had fallen. 
Another instance still more striking has fallen under my n0- 
tice. One mile from the foot of the hills, toward the bay, the 
county road is crossed by a small winter stream, never more 
- than six feet wide, which became perfectly dry early in August. 
The channel of the stream is not more than two feet in depth. 
In the latter part of August, I was surprised to find in it a small 
pool of water, at the side of the road. On the fourth of Sep- 
tember, the wet space extended some fifty feet. On the 9th of 
October, it had become a continuous, running stream, discharg- | 
ae five or six gallons per minute. And this happened without 
a drop of rain. 
nomenon. The water which falls in the winter and spring pea- 
to diminish the supply, and carry off entirely the water 
_ the bed of the stream in the intervals of the springs. 
_ Season advances, the days become shorter and power of the 
sun alsodiminishes. Evaporation becc tionally slowe?, 
cs i 
