226 Journal of Agricultural Research voi. xv, no. 4 



of the other biologic form had been present on the original material, 

 but was only given an opportunity to develop as a result of one or more 

 transfers to the common host. 



In experiments covering only a short period of time and with a small 

 number of forms it might easily be concluded that bridging had occurred. 

 The danger of drawing such conclusions, however, is clearly shown in 

 diagrams i to 4, inclusive. Many cases of apparent bridging were seen, 

 but in every case, with possibly one or two exceptions, it could be shown 

 conclusively that this was due to the fact that more than one biologic 

 form had been used. 



The cereals used were the following, unless otherwise specified: Oats, 

 Improved Ligowa (Minnesota 281); barley, Manchuria (Minnesota 

 105); wheat, Haynes Bluestem (Minnesota 169); rye, Swedish (Minne- 

 sota 2). Most of the grass seeds used were obtained by or through the 

 Minnesota Seed Laboratory. 



KEY TO TABLES I AND II AND DIAGRAMS I TO ID 



The results of experiments to determine the effect of bridging hosts on the para- 

 sitism of the rust forms are given in Tables I and II and diagrams i to 10. The results 

 of inoculations are usually given in the form of a fraction, the denominator indicating 

 the total number of leaves inoculated and the numerator the number which developed 

 uredinia. Whenever the precience of flecks is indicated, the number of leaves flecked 

 is given after the semicolon following the fraction. T ».o types of diagrams are used — 

 complete and condensed. Wheat, oats, barley, and rye are designated as W, O, B, 

 and R, respectively. The names of the grasses are either written out in full or the 

 key to the abbreviations is given in connection with each diagram. The sequence of 

 transfers is indicated by dashes, proceeding from left to right. All of the plants indi- 

 cated in the same vertical line after a dash were inoculated with the rust from the 

 host immediately preceding the dash. Not all of the inoculations are indicated, since 

 this would require altogether too much space. The essential ones, however, are 

 indicated in the so-called complete diagrams, while summaries only are given in the 

 condensed diagrams. 



In the condensed diagrams the small number immediately following the symbol 

 for cereal hosts or the names of the grass hosts indicates the number of urediniospore 

 generations on that particular host. The fraction in parentheses indicates the result 

 of inoculations which have been made during that period. The denominator gives 

 the total number of plants inoculated and the numerator the number which became 

 infected. The number of sets of inoculations is not indicated. Usually, although 

 not always, the immtme or highly resistant host was inoculated each urediniospore 

 generation. For instance, "R2 — B4 — Elymus canadensis^" indicates that two suc- 

 cessive transfers had been made to rye followed by foiu" successive transfers to barley 

 and five to Elymus canadensis. These are spoken of as lu^ediniospore generations. 

 The transfers were usually made at intervals of approximately two weeks, so that 



each urediniospore generation represents about tliat length of time. " R2 (wheat — ) — 



B4 (wheat — ) — Elymus canadensis^ (wheat — )" means that during the two genera- 



^5 35 



tions on rye 50 leaves of wheat were inoculated, none of which became infected. 

 One of the inoculations was probably made the first generation and the other the 

 second generation. The rust was then transferred to barley and kept there for four 



