576 MINNESOTA BOTANICAL STUDIES. 



The first samples used were circular, 39.7 mm. in diameter,, 

 and later were increased to 48 mm. All samples were weighed 

 as quickly as possible after separation from the leaf and then 

 subjected to the heat of a water bath oven for two hours at 

 100°C. All samples were cut from the leaf at 7 a. m. and 



7 P. M. 



The tests were conducted in three series as follows: 



1st. A sample was taken from the leaf on one side of the 

 midrib in the evening, and the second from the opposite side 

 in the morning, and their dry weights compared. 



2nd. The first sample was taken in the morning and the sec- 

 ond from the same leaf in the evening of the same day. 



3rd. A portion of the leaf of the required area was covered 

 by means of thin plates of cork, covered with black paper in 

 the morning, and one sample was cut from this shaded area 

 and another from the other half of the leaf lamina on the even 

 ing of the same day. 



The data obtained from these series are arranged in Tables 

 I, II and III respectively, and may be expressed briefly as 

 follows : 



I. Fifteen pairs of samples, taken between July 21 and Au- 

 gust 1, showed an average loss in dry weight during the night 

 amounting to 1.41 grams per square meter. In three of the 

 tests an increase was noted. (See Table I.) 



II. Ten pairs of samples, taken between July 27 and Au- 

 gust 8, showed a gain in dry weight during the daytime 

 amounting to 1.9 gram per square meter. (See Table II.) 



III. Ten pairs of samples, obtained from July 26 to August 

 8, showed that the average gain in weight of the unshaded 

 was greater than that of the shaded samples, in seven, and 

 less in three of the tests. The gain in dry weight of the un- 

 shaded samples was at the rate of 1.44 grams per square meter. 

 (See Table III.) 



