226 The Plant World. 



a and b. From this table it is seen that, according to the criter- 

 ion of total growth in length, the ranking of the different cul- 

 tures was II, III, V, I, IV. Thus cultures II and III were best 

 in both general appearance and in growth. The cause of this 

 may lie in the fact that air was not entirely excluded here, as in 

 I, nor was the water loss by evaporation from soil and pot so 

 great as in IV and V. Very little more water was absorbed by 

 II than by IV, but the total growth in length in IV was only 

 about 70 per cent, of that in II. Where the top of the pot was 

 unsealed the dry soil mulch had a depth of about one centimeter. 

 In these cultures the roots seem not to have suffered from lack of 

 air and to have had a sufficient supply of moisture. From this 

 result it rr.ight be concluded that Vicia faba succeeds as well or 

 better in an impervious container, when watered by this device, 

 as when grown by the method in common use. Further experi- 

 mentation would be necessary, however, to establish this conclu- 

 sion. 



The root system.s were exam_ined when the cultures were dis- 

 continued, and marked differences were noted in relation to the 

 m.ethod of watering. In those cases where the water had been 

 automatically supplied, the form and distribution of the roots 

 approached very nearh^ that which would have been exhibited 

 in the open soil. The roots were more branched and much more 

 uniformly distributed through the soil mass than in the control 

 culture. In the latter, a large proportion of the roots were found 

 massed against the walls of the pot, as is always the case with 

 pot-grown plants watered in the ordinary way, and a poor dis- 

 tribution of absorbing rootlets was observed in the interior of the 

 soil mass. There was practically no massing of the roots on the 

 surface of the pots in the cultures with the automatic arrange- 

 ment. It is thus suggested that an impervious container with 

 an automatic watering device m.ight be used in pot tests of soils 

 where it is essential that the root system be uniformly distributed 

 and uniformly in contact with the soil. In the present experi- 

 ments as striking a difference in root behavior was obtained, be- 

 tween the self-watered and the ordinary pot, as that obtained 

 by Livingston et al * and Reed f between the Whitney paraffined 



♦Livingston, B. E., Britton, J. C, and Reid, F. R., Studies on the properties of an unpro- 

 ductive soil. Bull. 28. XJ. S. Bureau of Soils. Washington. 190S. 

 Livingston, B. E., Paraffined wire pots for soil cultures. Plant World, 9: 62-66. 1906. 



tReed, H. S., The effect of certain chemical agents upon the transpiration and growth of 

 wheat seedlings. Bot. Gaz., 49: 81-109. 1910. 



