Mat 31, 1907] 



SCIENCE 



873 



up to optimum moisture condition. The 

 water content was kept up subsequently by 

 frequent watering. These pots, along with 

 two controls, were put into the greenhouse and 

 left standing for about two weeks before 

 planting to wheat in order to enable the treo 

 roots to become established in the soil. At 

 the end of this time, all pots were planted to 

 wheat, putting the same number of germi- 

 nated wheat seeds in each pot. 



The j&Tst crop of wheat was cut down at the 

 end of about three weeks and weighed and 

 the pots immediately replanted without dis- 

 turbing the soil. The wheat was similarly 

 planted and harvested every two or three weeks 

 until the middle of December, by which time 

 nine crops had been removed. In each crop 

 the average green weight of the plants in the 

 controls was considered 100, and the relative 

 weights of the others calculated on this basis. 

 The accompanying table shows the tabulated 

 results of the successive crops. There is 

 plainly a decrease in the green weight of the 

 plants grown in the pots with the trees. This 

 can not be due in any way to shade, as the 

 seedling trees were not large enough to inter- 

 fere in this way, and the pots were all 

 arranged in a row, so all had an equal amount 

 of light. The water content can not be a 

 factor, either, as all were watered every day 

 or two during the hot summer and every three 

 or four days during the cooler autumn. The 

 'plant food' removed can hardly be con- 

 sidered a serious factor in the case of such 

 small seedlings, especially as the crops in- 

 crease again, as will be pointed out below, and 

 there were of course no leaf washings from the 

 trees to affect the wheat. 



It seems, therefore, that the presence of the 

 roots must have had some other effect on the 

 growth of the wheat, as the size of the pots 

 made it necessary for the two kinds of roots 

 to be in close physical relation. That the 

 retarding effect is due to substances excreted 

 by the tree roots seems probable, and a closer 

 inspection of the table shows an evident in- 

 crease in yield toward autumn when the 

 physiological activities of the trees were 

 diminished by their entering upon their sea- 

 sonal rest. It was also noticed that the tree 

 pots that produced as much wheat growth in 

 November as the controls were the ones in 

 which the trees showed the earliest signs of 

 winter rest. Attention should be called to 

 the fact that if the tree seedlings removed 

 sufficient plant food to starve the wheat plants 

 in the summer period, the increase in yields 

 toward autumn would hardly be looked for. 



The increase in wheat growth in the various 

 pots toward autumn is more clearly brought 

 out in the last two columns. The first shows 

 the average of the preceding six crops, which 

 brings the time up to the middle of October, 

 about the time the physiological activities of 

 the trees would be decreasing, as the crop 

 harvested October 13 was not planted until 

 about September 20. The second of these 

 columns shows the average of the last three 

 crops. 



It will be seen that there is a decided 

 increase in the average yield of these three 

 crops over the average of the preceding simi- 

 mer crops, except in the case of the dogwoods, 

 and they were, excepting the pine, the last to 

 drop their leaves, having only dropped them 



