108 Messrs Wood and Berry, Selection of seed 



Smithfield yellow globes, which are really seedsmen's strains of the 

 same variety, all agree closely in high yield per acre, and low per- 

 centage of dry matter. Again the two intermediates show the same 

 concordance, as do also the two long reds, and the five golden or 

 yellow fleshed globes. Such agreement between these obviously 

 very closely related strains of the same varieties is very satisfactory 

 as indicating the reliability of the general results. Crimson 

 tankard appears to be quite a distinct variety, but neither its 

 yield nor its composition is good. 



Carter's sugar mangel, though a poor yielder, has such a large 

 percentage of dry matter that it stands fourth in order of merit 

 for yield of dry matter per acre. This variety is a hybrid 

 between the long red mangel and the sugar-beet, and in its 

 present condition possesses no particularly good points except its 

 high content of dry matter and sugar. 



It is by no means true to type, many pure white roots and red 

 roots occurring as well as pink ones, it has a very pronounced 

 tendency to run to seed, and a much-fanged root which buries 

 itself deeply, entailing much trouble in lifting. It is neverthe- 

 less an interesting variety on account of its high content of food 

 material, and is no doubt capable of improvement by judicious 

 selection. Carter's 1901 is also a very interesting variety. It is 

 stated to be the result of some years' selection, which has 

 apparently been carried out by growing seed only from roots of 

 high specific gravity. Its yield and composition are about on a 

 par with those of the golden or yellow fleshed globes, which it 

 resembles in other respects, but it does not grow so true to type 

 as they do. Turning to the diagrams expressing the results at 

 the other stations it will be seen that they confirm in almost every 

 case those at the University Farm, and this confirmation lends 

 further confidence to the results. 



It is now possible to classify the varieties which have been 

 submitted to experiment as follows : 



White fleshed yellow globes. Yield good, but quality low. 



Golden or yellow fleshed globes. Yield moderate, but quality 

 much better. 



Intermediates. Yield very fair, quality fair. 



Long red. Quality good, but troublesome to lift and only 

 suited to deep soils, on which it would have given a much better 

 yield. 



It was obvious early in the season that the yellow fleshed 

 globes were among the best varieties, but roots for individual 

 analysis for selection of mothers had to be saved before the com- 

 plete results for all the varieties were ready. Consequently the 

 varieties chosen for chemical selection, namely Sutton's golden 



