160 EXPERIMENT STATION. [Jan. 



erally considered tliat the older a lettuce soil, the better it is for 

 this crop, but if commercial fertilizers are employed indiscrimi- 

 nately in a lettuce house already well supplied with plant food, 

 the chances are that a case of malnutrition will result. 



Koses, carnations and violets require a rich soil, and a con- 

 siderable amount of manure is used by floriculturists in their 

 soil. Cases of malnutrition are prevented here by never grow- 

 ing these crops in the same soil more than one year, the benches 

 being refilled with fresh soil each year. A typical rose soil is 

 composed of one-third loam, one-third pulverized sod and one- 

 third cow manure. In addition to this, the plants are watered 

 once a week with a strong liquid manure. Cases of malnutri- 

 tion with this treatment seldom if ever occur with roses. 



A few years ago an experiment was conducted in one of our 

 houses devoted exclusively to the growing of American Beauty 

 roses. The soil was prepared as described above, and liquid 

 manure was applied freely once a week or oftener. The first 

 year the roses did well, and for the purpose of experiment we 

 attempted to grow a new crop of roses in the same soil which 

 had been used the previous year. The soil was partially ren- 

 ovated by the addition of new sod and some cow manure, and 

 besides this it received its customary application of liquid cow 

 manure. The plants had not been in the soil many weeks, how- 

 ever, before they commenced to die, and it was not unusual for 

 a number to die in a single week. The results of this experi- 

 ment were only what was expected, but a careful examination of 

 the plants was made which showed them to be free from patho- 

 genic organisms. The roots, however, were in a bad state, their 

 conditiou showing plainly what was the matter. Since it was 

 thought that this experiment had then proceeded far enough, 

 we decided to flood the beds with water, and make analyses of 

 the percolate which came through the bottom of the beds. The 

 beds were flooded for two hours each, and the water that came 

 through fli-st was, as might naturally be expected, highly colored, 

 Avhile that which came through later was clearer. The last 

 ])ercolate, after two hours' drenching, was remarkably clear. 

 Samj)les of this watcu* were collected at intervals of every fifteen 

 miimlcs. and chemical tests for acids and other substances were 

 made. The residts of the analyses were quite surprising, and 



