4 NATIVE LEGUMES OF NEBRASKA AND KANSAS. 



was therefore necessary to search out the grass fields that had been 

 pastured the least and which at the same time represented as nearly 

 as possible the average farm lands of the region. 



The next difficulty arose in the fact that the growing season of the 

 different species is very different. Some species of the genus Astra- 

 galus, for example, are in fruit by May and often dead before July. 

 Lotus does not come up until late, and the Psoraleas do not all show 

 their sprouts until about the middle or end of May, and by the first of 

 August they begin to break off just below the surface of the soil and to 

 blow away. Other species are not readily recognizable till late in the 

 season. From this it will be seen that anyone making counts will sel- 

 dom find all the legumes at any one time. 



A third difficulty presented itself in the irregular distribution of the 

 plants. It was not easy to find areas that represented average con- 

 ditions. Not only was there a natural source of error here but a per- 

 sonal one also. Having selected a piece of land that fairl}^ represented 

 a certain soil and slope it was then necessary to select the plots to be 

 counted. Here the personal equation appeared. With the amount of 

 time available it was not feasible to count large areas, so the square 

 yard was selected as the unit. If a patch contains showy legumes one 

 is almost sure to select plots that either have none of these plants or 

 have an unusual number. To avoid this unintentional selection the 

 following methods were employed : With eyes closed the writer would 

 walk a certain number of steps which would take him to a point he 

 knew he had not seen. For example, he determined to walk 100 paces 

 south and take his hundredth heel mark as the northwest corner of his 

 first square yard, or he walked backwards to a point he had not seen 

 and marked the plot to be counted in the same way. Either of these 

 methods gave him square 3'ards selected purely at random. It was 

 ;not practical to use the English method of throwing a hoop, and even 

 if this could have been done it is doubtful if any advantage would have 

 been gained thereby. 



With many legumes, especially Amorpha (shoe-strmg), Kuhnistera 

 (prairie clover), and several of the Psoraleas (in some places called 

 wild alfalfa), it is often impossible to tell whether there is one large 

 plant or several plants grouped together. In cases where there was 

 much doubt the group was always considered as a single plant. 



For the reason stated it is plain that the counts given in the follow- 

 ing table must be under rather than over the real number of legumes 

 present on the plots counted. It should also be noted that several of 

 the counts were made before the legumes had all come up. It was 

 hoped that several times as much data could be collected during the 

 3^ear, but the difficulties mentioned, together with the fact that much 

 of the season had to be spent in the semiarid region, limited the work 

 greatly. 



[Cir. 31 J 



