STUDY OF MORE ANCIENT RAINFALL RECORDS.

The period of actual gagings of the river flow at Portage having been so brief, it becomes important to learn how the rainfall in these years compared in quantity precipitated with the normal rainfall of the locality, and it is also important to confirm the accuracy of these river flow gagings by comparison of the yield per square mile on this watershed with that found on other watersheds, making due allowance for differences in rainfall.

A search for all reliable rainfall records in this region for a long term of years, finds only the following. These have been grouped according to

nearness to upper or to lower watershed and according to elevation.

It will be noted in comparing the above, that the rainfall in the plateau region tributary to the Genesee above Portage is distinctly larger than for the lower river; giving the stations equal weight, the average excess is 10.25 inches, and we may therefore expect the yield per square mile for the portion above Portage to be distinctly larger than for the entire area tributary above Rochester.

This is discussed more fully with the aid of compensated rainfalls and isohyetal lines upon a later page. (See page 188 below.)

All of these annual rainfalls have been plotted year by year on two diagrams, plate 17. On the first the ordinates are proportional to the actual precipitation; on the second the ordinates give the percentage which the

precipitation in each particular year bears to the normal or long term average for that particular locality.

The second diagram gives the most convenient comparison, and it is of particular interest to note how concurrent these records are, showing that the conditions are widespread and the precipitation substantially uniform over the watershed in question, and that a dry year in one part of this drainage is a dry year over it all.

This fact enables us to accept with more confidence the record of the fewer stations of the earlier years as representative of the whole territory now under consideration.

From these diagrams we confirm the conclusion already stated, that the year 1895 presented conditions of small rainfall and therefore of small runoff, severe as any that have occurred within the past half century.

From these diagrams we have sought to compensate the averages of short term rainfall observations so as to correct for the period in which they were taken having happened to be extra wet or extra dry, and the first step for this is the determining of the ratio of the precipitation to normal in any particular year. The method of compensation will be plain from the following table.

From the average rainfall at the many different stations as obtained in the foregoing table an isohyetal map has been prepared, a reduced copy of which is presented in plate 19.

The lines of equal precipitation are not so regular or so completely defined as we might like, but they prove most clearly the large excess of rainfall over the watershed above Portage in comparison with that on the lower Genesee.

These curves show the mean annual precipitation on the drainage area above Portage is 38 inches, while for the entire Genesee drainage above Rochester it is about 34 inches; and it would have been of some interest to compensate and increase that part of our mass curve that was based on Rochester gagings at least 13 per cent. so as to allow for that, and also to increase the portion derived from the Oatka Creek gagings, where normal rainfall is apparently about 33 inches, by 15 per cent.; but it has hardly seemed worth while to retrace these curves, particularly as they are on the safe side. The rate of increase in run-off would doubtless be larger than the percentage of increase in rainfall just noted.

Comparing the yield of the Genesee watershed with that from the upper Hudson watershed, we find the rainfall and run-off for the Genesee are much smaller; indeed the probable run-off year by year from the Genesee above Portage is just about one-half as great as that from the Sacandaga, notwithstanding the drainage areas happen to be substantially equal.

On the Sacandaga we have abundant forest cover and on the Genesee there appears to be a remarkably small amount; at least this is the case in the vicinity of the reservoir. No special tour of reconnaissance thrrough the extreme upper Genesee watershed for the purpose of judging of its quality has yet been made by a member of our staff other than that for studying the sites for small reservoirs, but so far as examined the drainage area was found nearly deforested.

QUANTITY OF WATER TO BE RESERVED FOR SCENIC EFFECT IN GLEN IRIS.

I have been very desirous of obtaining a series of observations at different stages of the river, each set consisting of an accurate measurement of the quantity of water flowing, combined with good instantaneous photographs showing the appearance of each of the three cascades with precisely this quantity of water passing, but conditions thus far have not been favorable.

These observations would tell better than any written description just what flow was reasonably necessary to be maintained for scenic effect. It was estimated by engineers of the City Engineer's office at Rochester from depth measured in the Portageville dam, that on the occasion of the visit of the Mayor and other city officials and of the State Water Supply Commission upon September 13, 1907, the flow was at the rate of about 100 cubic feet per second, but the conditions for a precise measurement at that time were not good. It is certain that the falls at that time presented a scene of great beauty and that the volume of water was not large.

During the period covered by Mr. Rafter's gagings in 1893 and 1896, the summer flow was for long periods only about 100 cubic feet per second, and during the season when this beautiful spot will be most frequented by those