46 REGENERATION 



hours in an oven at about 100°C. The result was as follows: 

 The dry weight of the 5 large stems (Fig. 37) was 13.670 grams, 

 and the dry weight of their 16 shoots was 0.495 gram. The 

 shoot production was therefore 36 milligrams per gram of stem 

 (all measured in dry weight). The dry weight of the 10 short 

 pieces of stem with 1 node each (Fig. 38) was 2.880 grams, and the 

 dry weight of 19 shoots was 0.115 gram, or 1 gram of dry weight of 

 stem produced 40 milligrams of dry weight of leaves. These two 

 figures, 40 milligrams and 36 milligrams, agree sufficiently closely 

 to show that under equal conditions the production of shoots of 

 defoliated pieces of stem occurs in approximate proportion with 

 the mass of the piece of (defoliated) stem; or, in other words, the 

 mass of shoots produced at the apex of the large defoliated stems 

 of Fig. 37 is approximately equal to the mass of shoots the same 

 stems would have produced had all the dormant buds of each stem 

 been able to grow out. 



The experiment in Fig. 35, where the stems were suspended 

 entirely in moist air, gave a similar result. The experiment 

 lasted from Oct. 4 to Nov. 7. Five large stems with 4 nodes each 

 (Fig. 35) having a dry weight of 5.486 grams produced 10 shoots 

 with a dry weight of 0.114 gram; i.e., 20.8 milligrams of shoot 

 per gram of stem. 



Four short pieces of stem with 2 nodes each, having a dry 

 weight of 3.214 grams, produced 8 shoots with a dry weight of 

 0.0668 gram; i.e., 20.7 milhgrams of shoot per gram of stem. 



A third stem was cut into 9 pieces with 1 node each (Fig. 36) 

 possessing a dry weight of 3.270 grams, giving rise to 17 shoots 

 with a dry weight of 0.050 gram; i.e., 15.3 milhgrams of shoot per 

 gram of stem. 



The first two figures are identical, the last figure is a httle low. 

 In these experiments the end of the piece may suffer (by drying 

 out or falling a prey to fungi) and this creates an error which is 

 especially noticeable when a stem is cut into many small pieces. 

 But in spite of these sources of error the results are remarkaljly 

 clear and consistent. 



It seemed of interest to compare the behavior of defoliated 

 stems split longitudinally. In this case the two halves should 

 give approximately equal results. 



2, Experiments with Stems Split Longitudinally. — Experi- 

 ments were made with stems split longitudinally as indicated 

 in Fig. 39. Only pieces from the middle of the stem of a large 



