278 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



measure due to the thorough preparation of the soil and to con- 

 stant attention to cultural details during the growing period. The 

 cultivation of the plants was under the charge of the Superintendent 

 and the garden foreman, Mr. J. Wilson. 



In the seedling stage some of the plants suffered an attack from root 

 maggot, which is a common pest at Wisley, and for which no remedy 

 has been found. Happily the plants quickly grew out of this trouble. 



Inasmuch as cross-fertilization does not occur in the pea, except 

 in very exceptional circumstances, it was expected that the stocks 

 would prove true to character. Although this was generally the case, 

 there were, in a few stocks, a surprisingly high proportion of rogues. 

 The most striking among the rogues were, of course, the " tare-leaved " 

 forms, which only persistent roguing on the part of growers can keep 

 under, and which appear to be constantly occurring among the ordinary 

 stocks of peas. As is well known, the " tare-leaved " rogue pea is 

 characterized by the leaf tips being pointed and not flattened, and 

 by the possession of curled pods which bear peas of a peculiar and 

 distinctive flavour. It does not, however, appear to be so well known 

 that " tare-leaved " rogues exist not only in this well-defined form, 

 but also as what may be called " incipient rogues," and that these 

 " incipient rogues " give rise in the following generation, as Professor 

 Bateson has shown, to thorough-going rogues. Hence in the roguing 

 of peas it is important to eliminate the incipient as well as the well- 

 marked tare-leaved forms (fig. 88) . 



A second class of rogues, which occurred in a large number of stocks, 

 consisted of tall-growing forms which appear among dwarfs or semi- 

 dwarfs. The present state of our knowledge indicates that the 

 presence of such tall individuals is in all probability due to careless- 

 ness in some operation connected with the harvesting of the seeds. 



A point which is worth bringing to the notice of growers is 

 the frequency with which one-podded forms appear among stocks 

 containing two-podded forms. Selection would eliminate the one- 

 podded forms, but of course that must mean the rebuilding of the 

 stocks from a single, uniformly two-podded plant. In this connexion 

 it may be pointed out that not all plants which are fairly uniformly 

 two-podded breed true to this character ; on the other hand, some 

 undoubtedly do so. It would, therefore, be necessary in building 

 up a two-podded stock to start with several two-podded parents and 

 keep the progeny of each separate until it had been ascertained in 

 the next generation which was breeding true to two-poddedness. 



Classification. — The system of classification adopted in the Table 

 is as follows : — The stocks are first grouped into first- and second-early 

 varieties. These groups are divided into four classes according to the 

 height of the variety, and each class is subdivided according to seed- 

 form (round or wrinkled). In this subdivision varieties with " dent ' 

 or slightly wrinkled seeds are classed with the round. The varieties are 

 arranged alphabetically under their seed character. 



The pea which proved earliest, viz. No. 31, although it failed to 



