20 METEOROLOGY. 



mountains. The difference which would be expected from this rate of decrease of temperature 

 is 9°. 3, showing that, even when the wind blows from the coldest quarter, less cold was 

 produced than is usual in places having the same relative altitude above the sea. And as the 

 westerly winds are most prevalent, the difference between the climate of the pass and Bitter 

 Root valley must usually be even much less. 



It has been foimd that some snow usually exists in the Nachess'Pass from November to May; 

 but that being 1,424 feet higher than the Snoqualme, a month may safely be allowed as the 

 difference in the Snoqualme Pass, while the period during which more than five feet lies in the 

 highest part of it (ten miles, over 2,000 feet) may be reduced to a month. 



In calculating the probable amount of snow which fell in the Snoqualme Pass in the winter 

 of 1853-'54, the mean temperature of that winter should be used instead of the mean oi four 

 winters, since there is a marked difference between them — that of 1853-54 being 1°tVo colder 

 than the average, and i°j% colder than two winters in six since 1849, and only IVA warmer 

 than the coldest. 



Using the recorded temperature at Fort Steilacoom for 1853-54, we find that November, 

 1852, was only 0°yVo higher than that of December; therefore, if snow fell in December it 

 must also have fallen in November. And in this month there was 18yVo inches of rain at 

 Steilacoom, while only 20iW fell in the three following months. This would give for three 

 months preceding Mr. Tinkham's journey across the pass in Febniary 26iYo. (adding Novem- 

 ber to Captain Humphrey's estimate,) and 12i-Vo afterwards. By the rule of allowing twelve 

 times the bulk of the rain for snow, Mr. Tinkham should have found twenty-six feet, whereas 

 he found only six; showing that either 7yV inches only of moisture had fallen, or that most of 

 it (over three-fourths) fell as rain. Admitting that the whole which fell after his crossing was 

 in the form of snow, there would be 12yVofeet more; but this would be absurd, since February 

 was nearly 10° warmer than January, and there was 1 yVg- less rain at Steilacoom ; taking the 

 same proportion as for the preceding three months, viz: y^,, making three feet to be added — in 

 all, nine feet for the winter. But as the temperature of March was 0°y?-oV colder than December, 

 we must admit that two-fifths of the moisture of that month was also deposited as snow; and 

 then 2yVV inches of rain at Steilacoom would give us + yVo feet more of snow to be added, 

 making 9-iVo feet from November 1 to April 1. 



If November is omitted, so must be December ; and then the moisture which fell before the 

 21st of January at Steilacoom, 3yVo inches + 12, would give 3yVo feet of snow, while Mr. 

 Tinkham found six feet. Therefore, November and December must be included. Besides, 

 Lieutenant Mowry states that snow falls in these mountains in November, and Lieutenant 

 Hodges actually met with a slight fall in the Nachess Pass in September. 



Here we see, too, the necessity of taking into the account the loss by evaporation, thawing, 

 and condensation, which goes on even in the coldest weather of the arctic winter. 



Supposing this to account for the difference of the calculated and observed fall from Novem- 

 ber 1 to January 21, and admitting the fall from January 21 to April 1 to be three feet, added 

 to the six feet found by Mr. Tinkham, there would remain on the ground at the latter date 

 only two feet. This would be much further decreased by the greater amount of thawing in the 

 two warmer months following January. 



Or, by another process, we find that the difference between 39 feet, (26 + 13,) the amount 

 calculated by the rains of Steilacoom and the true amount, would be in the same proportion as 

 that between the amount predicated on the rains previous to January 21 and the actual depth 



