EXPERIMENTS W ITH PERMANENT PASTURES 17 



or 6 applications of fertilizer in the 10-year period shows that the series which 

 received 5 applications slightly outyielded those which received 4 and 6, 4 being 

 the lowest of the three. It is believed that these differences are too small to be 

 of any significance. It appears that under the conditions which prevailed during 

 this experiment, drought and all considered, 4 applications of fertilizer furnished 

 an abundance of nutrients — that is to say, a sufficient quantity of minerals ap- 

 parently had been added. Nitrogen was a serious limiting factor in this experi- 

 ment, and that, if supplied in a soluble form, should be added annually. 



In table 11 is given a summary of the yield data arranged according to fer- 

 tihzer treatment, with and without lime. An examination of these figures shows 

 that marked benefits resulted only from lime and from the nitrogen-phosphorus- 

 potassium combination. By comparing table 11 with table 8 it may be seen that 

 the marked response to minerals obtained in 1923 did not manifest itself in 1930. 

 The reason for this difference apparently lies in the proportion of white clover in 

 the vegetation of the two years. As Haskell (7) has pointed out, white clover 

 "comes and goes." Irrespective of the soil treatment, there is a pronounced varia- 

 tion in the proportion of white clover from year to year. In so-called clover years 

 white clover appears abundantly on soils supplied with plant nutrients and 

 moisture, in off years it disappears from those same areas due to causes which 

 are not understood. 



1923 was a white-clover year, 1930 was not. In support of this point, note 

 the difference in the percentage of white clover in the plots of series I in 1929 and 

 1930, as given in table 12. This series had the highest percentage of clover of all 

 series in both years. Why the marked change in flora in the period of one year? 

 ^^'as the drought of 1929 an important factor? The causes of such changes are 

 not known, but the changes do occur and should be taken into consideration in 

 the interpretation of pasture data. Thus, it seems reasonable to conclude that 

 the lack of response to minerals in 1930 was due to the fact that white clover 

 was almost entirely absent from the plots in that year. On the other hand, the 

 response to nitrogen was marked because of the grasses. 



In table 13 is given the percentage of protein, phosphoric acid and calcium 

 oxide in the 1930 clippings. The figures show in a general way the agreement 

 between character of fertilizer applied and composition of the crop. Vegetation 

 from plots which received nitrogen, phosphorus, and lime was richer in those 

 elements than that from plots which did not receive the same elements in fer- 

 tilizers. The difference in respect to nitrogen (ex-pressed as protein) is more 

 marked in 1930 than in other years due, no doubt, to the practical absence of 

 clover in 1930. 



It is believed that the higher content of nutrients in vegetation from the 

 limed and fertilized plots accounts for the greater palatability of the vegetation 

 as noted in the field. This applies particularly to lime and nitrogen. It was 

 repeatedly observed that the plots which received these nutrients were more 

 closely grazed than those which did not. This can be seen by a close inspection 

 of figure 8. 



Experiment C. — This experiment was laid out in 1924 by M. O. Lanphear 

 in order to obtain certain field data to be used in an ecological study of pasture 

 cover. The area occupied by this experiment was particularly attractive for the 

 purpose on account of the poor quality of herbage. This, as shown by figure 9 



Formerly instructor in Dcpt. of Agronomy. Work ro])orted tiore was part of thesis submitted 

 for M. S. degree. 



