100 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



the water of Lake Geneva. These observations have been 

 systematically and carefully made for forty years, and in all 

 cases the bench marks to which the levels of the water are 

 referred are points so well known that the entire series of 

 observations can be considered as perfectly comparable 

 among themselves, except in so far as a slight uncertainty 

 attaches to a portion of the observations, due to an accident 

 which occurred to the recording apparatus. After carefully 

 correcting: and reducing the whole series of observations to 

 a uniform standard, it appears that the annual variation in 

 the height of the lake is distinctly pronounced for each year 

 of observation, but that the water returns to the same level 

 at the same epoch of each year. The lake appears to have 

 been lowest in December, 1857, and to have been highest in 

 July, 1846 ; the entire range between these figures being 

 only seven feet. It is, on the average, lowest on the 11th of 

 February, and highest on the 7th of August. The connec- 

 tion between the accidental or extraordinary variations in 

 the level of the lake, and the meteorological conditions, es- 

 pecially rain and melting snow, is fully examined ; and it 

 appears that the heat of the autumn is a very important 

 factor in determining the volume of the waters of the Rhone, 

 and consequently that of the lake. By means of a table 

 giving the quantity of rainfall, and the temperature of the 

 spring for each year as compared with the averages, Planta- 

 mour is able to explain away so large a portion of the ir- 

 regular annual variations of height of the water that the 

 remaining discordances are remarkably small; while there 

 are groups of positive discordances, which have suggested 

 the propriety of comparing the heights of the lake, year by 

 year, with each other, in order to deduce the secular varia- 

 tions, if any exist. He has, therefore, divided his series into 

 periods of nine years each, from which grouping it is shown 

 that the level of the maximum water has not varied in any 

 progressive manner in the course of thirty-six years, and that 

 the annual fluctuations depend upon the dryness or wetness, 

 the cold and the warmth of the year. If, however, within 

 each of these groups, we compare the levels of the water, w r e 

 find that the heights at low water arc larger in proportion 

 as the rainfall diminishes, and a gradual increase in the 

 level of low water is noted, which can not be explained by 



