E. W. Morley— Oxygen in the Ae. 175 
some samples were collected in clean stoppered and ca 
bottles, and kept for a short time by inverting the bottle in the 
cap which had been filled with water. In this case the air 
was withdrawn for analysis with a Tcepler’s mercur 
The table gives in the first column the date, in the second, 
the mean temperature of the day at this place as determined | 
by three observations. In the third, on the days when analyses 
were made, the hour of collecting the sample is given, frac- 
tions of hours being disregarded. In the last two columns are 
through a capillary tube twelve or twenty times. All made 
between the first and last dates of the table are given without 
selection, except that some were rejected for obvious instru- 
mental errors. 
e remarkable deficiency of oxygen observed on the 
twenty-sixth of February seems affected with no reason for 
d On Sept. 16, 1878, two very careful analyses of the 
same sample gave 20°49 and 20°46 per cent of oxygen. On 
of air from the Bay of Bengal showing 20°46 per cent, one of 
air from near Calcutta, showing 20°39 per cent, and one of air 
from near Algiers, showing 20°41 per cent. That Jolly and 
the writer have found air almost as deficient in oxygen as the 
three last will lessen the probability that the air of the surface 
