NUTRITION OF DROSSllA ROTUNDIFOLIA. 21 



the leaves, because many ordinary marsh-plants have equally 

 small roots. Cohn concludes that the fact of Dioncea, Sarracenia, 

 or Nepenthes thriving under culture when deprived of animal food 

 is in no way contradictory to the belief that the leaves can digest 

 nitrogenous materials. 



My experiments were conducted in the following manner. 

 The Drosera plants were obtained from a neighbouring common 

 on June 11, 1877, and were planted in moss in six ordinary soup- 

 plates. The plates were placed in two rows on a wood tray having 

 a raised border all round, and were covered by a wooden 

 frame 1 foot (about 30 centims.) in height, over which gauze 

 netting (with a mesh of 1/4 millim. diameter) was stretched. 

 This gauze was similar to that used by my father in his crossing 

 experiments, and known to be effective in excluding insects. The 

 frame lifted off and on like a bell-glass, and fitted close within 

 the rim of the tray. Hardly any insects penetrated into the case ; 

 but I did not particularly attend to this point, because any insects 

 caught by the starved plants could only render my results less 

 striking, but could introduce no error. The whole apparatus 



stood near the light in a grape-house where no artificial heat was 

 applied. The shade produced by the vine-branches and by the 

 gauze, appeared to suit the plants, as they throve wonderfully. 

 The plants and the moss in which they grew were kept very 

 moist ; and by frequently pouring out the water in the plates and 

 adding a fresh supply, the water was constantly renewed. 



Owing to a delay in beginning the experiment, the plants were 

 well grown when collected. The results might have been more 

 striking if the experiment could have been commenced with 

 younger plants ; but as it was, I had at least the advantage of 

 knowing that the plants were perfectly healthy. Each plate was 

 divided by eye into two halves separated from each other by a 

 thin piece of wood hardly reaching above the surface of the moss. 

 That half of each plate which appeared least flourishing was 

 selected to be the " fed " side, the opposite side being labelled 

 " starved " # . The plants grew so close together that it was difficult 

 to count them accurately ; but the following Table gives the 

 numbers as counted roughly. 



* For the sake of clearness in my notes &<\, I used the word " starred " in 

 place of the more correct "unfed." 







