VEGETATIVE TROPAGATION OF LEOUMTNOTTS FORAGE PLANTS. 85 



No losses resulted on one occasion in the transfer of l.SOO plants to 

 the permanent nursery rows 5 miles distant. The tops were cut 

 back to 6 inches in height before being removed from the pots in 

 the cold frames. (See PI. Ill, figs. 1 and 2.) 



APPLICATION OF THE METHOD TO PRACTICAL PLANT-BREEDING 



PROBLEMS. 



In connection with establishing new varieties of such leguminous 

 forage plants as alfalfa and clover it is sometimes desirable to start 

 with a strain from a single individual, or at best from a limited 

 number of individuals. This is the case where an especially prom- 

 ising form is confined to so few plants that the problem of increasing 

 the stock for further tests and possible introduction is a serious one. 

 In work of this kind many difficulties have heretofore been encoun- 

 tered. The seed selected from a promising set of individuals in an 

 ordinary nursery or testing plot ma}' have as its male ])arents plants 

 of all of the strains in the series mider test, a circumstance which 

 works against the fixing of the strain along the desired lines. 



This promiscuous parentage can usually be avoided only by keep- 

 ing the remainder of the plants clipped to prevent flowering. This 

 is not practicable in case other strains are being developed at the 

 same time. \Vliile it is possible to isolate several hundred plants so 

 that the danger of outside pollination is for the most part eliminated, 

 yet with a few plants this is much less satisfactory, as there is not 

 the protection of numbers which a considerable area of plants of one 

 strain gives. The several hundred plants which can readily be pro- 

 duced from the selected individuals during the winter can be isolated 

 by transfer to a considerable distance from other plants of the same 

 or closely related species. It is possible that a considerable area 

 could be practically isolated by lateral screens to confine the fer- 

 tilizing insects temporarily to the plants in question. Under these 

 conditions the presence of great numbers will make it probable that 

 the bulk of the seed secured wdll have the selected individuals for 

 its male parents. 



The quantity of seed procm'able from a few plants is usually so 

 small that several seasons are required to obtain sufficient stock for 

 even the preliminary tests of the new strain under field conditions. 

 But by using cuttings it is quite practicable to produce in the green- 

 house as many plants from one individual during the fu'st winter as 

 would be expected in at least two years from seed. Therefore this 

 method results in a considerable shortening of the time required to 

 get the seed of any one selected strain in sufficient quantities for 

 field tests. 



In practical selection work where strams resistant to cold, drought, 

 or disease are being developed, natural selection will weed out the 



102-lV 



