344 



were 2 in whieh the soil remained fallow, making the total number 79. 

 Of the 7 east and west rows, 1 was devoted toeaeh of tlie seven kinds 

 of plants. Of 11 north and south rows, 2 were unnianured, 2 received 

 the mixture of the three fertilizing materials, 2 nitrate of soda, 2 sul- 

 phate of potash, 2 plain superphosphate, and 1 double superphospliate 

 in 1885 and 1880 and plain superpliosphate Mith sulphate of potash in 

 1887. Each test A\ath each plant and fertilizer was thus made in 

 duplicate, except in the single row last mentioned. The same kind of 

 plant was grown in the same box year after year and with the same 

 fertilizer, except in the case of the single row referred to. 



To facihtate proi)er distribution of the pollen of the lui»ines, a hive 

 of bees was kept near the phuits. Instead of a cover of wire netting, 

 which is soioetimes used to prevent the ravages of sparrows Avhen the 

 seeds are maturing, a Geier (a kind of small hawk) was ke])t contined 

 by a chain close by, and for a short time strings carrying strips of colored 

 ])aper were stretched over the plants. No trouble from birds was 

 experienced, though the Geier have in other instances not sutliced to 

 fiijiliten the sparrows away. 



The observations on the plants harvested, which are rejjorted in 

 detail, include weight of whole crop and of seed and straw (including 

 chaff) separately; the number, length, and weight of stems; weight 

 of leaves and of ears or pods from which the seeds had been remo\"ed; 

 number and weight of ears or jiods with seed, and weight of seed iter 

 «'ar or pod. large and small; averages per plant for total weight, and 

 weiglitsof seeds, ears <»rj)ods, leaves, and stem separately; and calculated 

 yiehl of seed, straw, and wlude crop pti hectaie. The weights as given 

 refer to air dry material. Determinations were made of the percent- 

 ages of nitrogen, ash, potassium »»\ide. and ph(»spliorus pentoxide 

 in the whole i>rodnee and in tlie st-t'd and straw separately, and from 

 these ar*' made calculations of tlie amounts contained in the i)ro<l- 

 nce per ex])eriment and per heitare. No determinations were mafle 

 of tlie ann>nnts of moisture, so that the figures do not show the aetnal 

 comitosition of the plants, /. c. of water-free substance. The samples 

 analyzed were taken from the boxes of oidy 1 of the 2 duplicate rows of 

 }»laiits whieh received like treatnu'iit. Hence while the data foi- quan- 

 tity of air-dry material represent averages of two duplicate trials, those 

 for chemical composition represent only one of the trials. 



Barometric jnessure, rainfall as shown by rain gauge, temix'rature of 

 air, and t»inperature of soil at the surface and at ditlerent d»'i)tlis of the 

 soil ln'tween the boxes, ;us indicated by thenu<tmeters, were observed 

 jind ri'corded during the growing period from Ajuil t^) !Sept«'nd)er. 

 Determinations of evaporation of water by the method of H. Wild of 

 St. Pet«'rsburg were attemi»ted. but fouml nnsatisfaetory and given uj). 



For study of the water snp]»liedto the soil, determinations were made 

 of the amount of Mater, (1) in the soil at the beginning as found by 

 moistui'c determinations in a sami)lc; (2) supplied by raiu as measured 



