Orchard Cover Crops in Delaware. 207 



large proportion of the roots were within 8 inches of the surface. The tap root 

 was more or less branched, the branches going downward also, but at a depth 

 of about 3 feet the roots were dead and partially decayed. There was a sur- 

 prisingly small amount of tubercles on the roots ; a good many small ones but 

 few large ones. This was true of the hairy vetch and crimson clover also. 

 The opinion prevails among several members of the Station staff that the 

 limited amount of tubercle formation is due to the presence of an abundance of 

 nitrogen compounds in the soil, and that there is little necessity for the plants 

 getting nitrogen from the atmosphere by the aid of bacteria. 



The crimson clover was very shallow-rooted, the tap root usually extended 

 downward about one foot. The vetch was about as deep-rooted as the alfalfa. 



The observations by Craig, mentioned in Cornell Bulletin 108, that vetch 

 rootlets follow the burrows of earth worms was noticed here with both vetch 

 and alfalfa. 



SOIL MOISTURE IN PLOWED GROUND VERSUS UNPLOWED GROUND WITH DEAD OR LIVE 



COVER CROP ON. 



A full discussion of this question is given by Prof. Penny in Bulletin 60, 

 so only a brief mention of the subject will be made here. The cover crop plots 

 at the Experiment Station were long and narrow. A portion of the orchard was 

 plowed crossways of the long way of the plots May 15, 1902, and the remainder 

 was left unplowed so the cover crop might grow all summer. Each plot was 

 thus about two-thirds plowed and one-third unplowed. The soil samples for 

 the determination of the moisture content were taken from the plowed and un- 

 plowed portions at the same time at intervals of one week from May 19 to 

 July 29. The portion plowed was kept well cultivated throughout the season. 

 During May and the first 20 days of June there was a small amount of rainfall 

 and there was considerable difference in the percentages of moisture in the tilled 

 and unfilled soil. On the 21st, 26th and 29th of June heavy rains fell and the 

 difference diminished considerably. There was a heavy rainfall in July and 

 hy the end of the month there was slightly more moisture in unfilled portions 

 which had dead covers as rape, turnips, cowpeas and soy beans, than there 

 was in the tilled portions. Where there were live covers there was less 

 moisture in the uncultivated than in the cultivated portions July 29. 



Hairy vetch plot. — On May 19 there was a difference of 5 per cent of soil 

 moisture in favor of the tilled portion. From that date there was a steady 

 gain in moitsure in the cultivated ground until June 16, when there was a 

 difference of 10 per cent in its favor. Copious rains a few days later made the 

 amount of moisture about the same throughout the plot until July 29. 



Alfalfa plot. — There was a difference here in favor of the tilled part of from 

 31^ to 814 per cent of soil moisture for a month following May 19. Then 

 for two weeks the unfilled part pained in moisture ; during the next two 

 weeks it lost and then gained a little again the following two weeks. By 

 the end of July tlie tilled portion had 4 per cent more soil moisture than 

 had the portion upon which the crop was still growing. 



Crimson clover plot. — The cultivated portion contained from 41/, to 8 per 

 cent more moisture than the uncultivated portion from May 19 until the mid- 

 dle of June. After that there was little difference because of abundant rains. 



Red clover plot. — The difference in moisture in this plot between tilled and 

 unfilled portions varied from 41/4 per cent May 19 to 9% per cent June 16 in 

 favor of the tilled portion. The variation then decreased to 1 per cent June 

 30 ; then it increased to 714 per cent July 21, and fell rapidly to 1 per cent a 

 week later. 



Mammoth clover plot. — The record for this is similar to that of red clover, 

 but with less sudden fluctuations in the per cent of moisture in favor of the 

 tilled part. 



Rye plot. — There is less difference in favor of the tilled part than there is 

 in any of the other plots mentioned above. The differencs were 3 per cent 



