32 



Experiment Station Bulletin 345 



curve was the same as that for 1940. However, the soil on the cultivated 

 plots froze to a depth of 22 inches only, whereas the depth of frost in 

 the grass plot was about 1 2 inches. The frost this year as last came out 

 from both top and bottom, but during this period there was no rainfall 

 and, therefore, no soil loss. The frost was completely out on April 17. 



During the winter of 1941 a sampler for soils, combining some of 

 the designs of both Coile and Bradfleld, was developed. This sampler 

 is made so that a 3}£ inch x 6 l / 2 inch cardboard cylinder held by an 

 outer casing of steel can be driven into the ground to remove a core of 

 undisturbed soil. The sample of soil is then stored in the cardboard 

 container for future analysis. Such samples from the plots were sub- 

 jected to air space, percolation, and water stable aggregate analyses 

 during the year with the idea of using the data for statistical treatment. 



Fig. 3. A New Method of Driving and Dividing Core Samples of Soils 



The presence of a large number of earthworms has been noted on 

 the run-off plots, but no attempt to evaluate the numbers has been made 

 except to count the numbers which are found in the tanks following a 

 rain. Observations during the past season have indicated that earthworms 

 may be of more importance from the standpoint of soil erosion and yield 

 than is usually recognized. 



The relative numbers of earthworms may be estimated by the num- 

 ber of feces left on the surface of the soil following a long rain. Obser- 

 vations on the three fields the past fall showed that the number of earth- 

 worms was in about the same proportion as indicated by the per cent 

 of aggregates. 



Counts of the number of millipeds found in the tanks were made at 

 intervals during the past season. Again they caused no injury to the 

 potatoes which were grown on the plots. 



