4 BULLETIN 79, U. S. DEPARTMENT OF AGRICULTURE. 



In gathering the experimental material each year, a sufficient num- 

 ber of representative plants was selected to give about 200 ripe leaves 

 near the bases of the plants, six or eight leaves being taken from each 

 plant. The leaves were taken in pairs, each alternate leaf being placed 

 in one lot and the other leaves in a second lot, the aim being to have 

 each leaf in the one lot represented by a duplicate in the second lot. 

 For the success of the experiment it is essential that the total dry- 

 weight and also the average composition of the two lots of leaves 

 be practically the same. The detailed examination of the material 

 gathered in the manner described indicates that these requirements 

 have been satisfactorily met. As promptly as possible after harvest- 

 ing the leaves the midribs were completely removed from the lot 

 which was used to obtain the original dry weight. This was done to 

 prevent any flow of cell sap between the leaf web and the midrib 

 during the drying. The split leaves as well as the midribs were at 

 once placed in a large drying oven and maintained at a temperature 

 of about 80° C. until all were completely dry. The temperature em- 

 ployed rapidly kills the protoplasm, and comparative tests with other 

 methods of killing, such as plunging in boiling absolute alcohol, 

 showed that the respiration changes during the drying were too small 

 to be of significance for our purposes. (See p. 35.) The dry weights of 

 the leaf web and the midribs were obtained and the material preserved 

 for analysis. The second lot of leaves was placed in the barn for 

 curing in the usual way without removing the midribs. When the 

 cured leaves were ready for examination the leaf web was removed 

 from the midrib, the dry weights obtained, and the material preserved 

 for analysis. 



As has been stated, observations were made for a period of four 

 consecutive years of the loss of dry matter in curing when the leaves 

 are harvested by picking them from the stalk. The detailed results 

 of these experiments are shown in section A of Table I (p. 5). In the 

 case of the 1908 material the curing was stopped before the midribs 

 had cured, but the curing of the leaf web was practically complete. 

 In 1910 a special test of partial curing was made (Table I, fourth 

 column) and in this case the curing was stopped while the larger veins 

 and midribs were still uncured. In 1911 a test was made of the effect 

 of artificial heat on the curing. 



