1912. 



PUBLIC DOCUMENT — No. 31. 



61 



finer particles, clay, etc., tliaii is generally found here in our 

 soils. In our experiments we selected soils which resembled as 

 much as possible the best types. Some of the soils in the Con- 

 necticut valley seemed to be excellent substitutes for the real 

 rose soils, and these were tested in order to obtain some idea of 

 their relative value. Most of the Connecticut valley soils run 

 low in coarser materials and high in the finer materials such 

 as very fine sand, and occasionally silts predominate. These 

 fine sands and silts give a compact character to the soil, with 

 a large water retaining capacity. Very fine sand predominates 

 in many of these soils, as high as 75 per cent, being found in 

 some samples, but nearer the river we often find samples in 

 which the coarser silt predominates. 



These experiments were continued for two years in a green- 

 house provided with 5 beds, one-half of each bed containing a 

 typical Amherst loam which had been used for some years in 

 growing cucumbers, tomatoes and melons, and to which no 

 commercial fertilizer had been added, although horse manure 

 had been applied freely. The other half of each bed was filled 

 wdth specially selected soil of different textures for the pur- 

 pose of comparison. These benches were 1 foot deep and the 

 bottoms were filled with brick placed on end, leaving about Y 

 inches depth of soil in each bed. Seventeen plants were used 

 in each section, or in the whole house a total of 170 plants. 

 Since five of these sections contained soil of the same types, 

 only five were used for other types, and these were all different. 



Table 1. — Showing Mechanical Analysis of Soils used in Rose Experi- 

 ments. 



