COLORADO IRRIGATION WATERS AND THEIR CHANGES. 65 
a month of low water, carried only 10.8 grains. This 
flood water, after storage for twenty-two months, carried 72 
grains. Our data is not sufficiently full to enable us to go further 
with this discussion. 
§ 119. The work done by the irrigating waters of the Ar¬ 
kansas valley is evidently similar to that done by them in the 
Poudre valley, and so far as the rate at which the salts are carried 
out of the soil is indicated by the contents of total solids per gal¬ 
lon, it is very nearly the same, differing principally in carrying a 
fairly large percentage of sodic sulfate, while the Poudre return 
water carries relatively but little or none of this salt. 
THE SUSPENDED MATTER CARRIED IN TIMES OF HIGH WATER. 
§ 120. This will vary both in amount and character, accord¬ 
ing to the conditions prevailing within the drainage area of the 
streams carrying it. The Rio Grande, in New Mexico, would 
scarcely be expected to carry the same character of suspended 
matter, especially after a torrential rain somewhere within its 
plains section, as at Del Norte, Colorado, after a similar rain in 
the mountain districts to the west of it. 
§ 121. The amount of sediment, as I have found it, has been 
a great disappointment to me, it being very small in amount com¬ 
pared with my preconceived notions, and of a somewhat different 
quality. 
§ 122. On May 22, 1902, we had an excellent opportunity 
of obtaining a sample of Poudre flood water, caused by a heavy 
rainfall within the foothills, whereby the river was swollen to 
such an extent that it passed beyond its bounds. It carried on 
this date 12,000 second-feet, or about ten times its usual volume 
at this season of the year. This water was very thick with mud 
and debris, such as the unusual volume of water would tear loose 
along its course. I had a large sample, 102 pounds, of this water 
collected from the middle of the stream. The bucket with which 
the water was dipped was allowed to sink as far as it would in such 
a current. The whole sample was allowed to settle for several 
days, on account of the suspended clay, and then filtered. The 
suspended matter amounted to 0.213 per cent, or 2,130 parts per 
million. The analysis of the sediment gave the following results: 
