EXPERIMENT STATION RECORD. 



Vol. VI. No. 6. 



Garden herbaria, or collections of specimens illustrating the species 

 and varieties of cultivated plants, have been begun at a few of the 

 experiment stations, notably under direction of Prof. L. H. Bailey, at 

 the station connected with Cornell University. The object of these 

 herbaria is to furnish a scientific basis for the study of varieties by 

 providing a fund of exact information for their identification, and, 

 what is of vastly greater importance, for the guidance of experiment- 

 ers who are seeking to improve them. Such collections help to show 

 the relationships of varieties and indicate features to be elaborated 

 and improved, and others to be discarded. It is an acknowledged and 

 regretted fact that in the study and comparison of different varieties 

 of plants, and particularly of fruits and vegetables, a decided draw- 

 back sometimes exists in the absence of accurately determined speci- 

 mens, and even of precise descriptions or figures. It may occur that 

 some of the varieties considered have been misidentified, or that the 

 plants under observation vary more or less from the original stock, and 

 in various ways uomenclatural and other errors may inadvertently 

 creep in, affecting the value of the results reached, and possibly ren- 

 dering them absolutely misleading. A remedy for such a condition of 

 things exists in the garden herbaria, which have been found almost 

 invaluable aids in varietal -work. 



The scope of these collections may vary, but should be as extensive 

 as practicable. They should primarily comprise dried specimens of 

 the different varieties cultivated, and in the case of annuals should 

 be enlarged each year by the addition of plants grown from seed. Not 

 only should tyjiical representatives of the varieties be shown, but also 

 marked variations should be preserved. Where the whole plant can 

 not be pressed, leaves from different portions and sections of the stem 

 and branches may be selected. In the case of fruit trees, for instance, 

 leaves should be saved from new shoots of the year's growth, and also 

 from fruiting branches. The blossoms, in bud and full bloom, should 

 be carefully pressed, so as to retain the relations of the various parts. 

 Small packets of the seed, both green aud ripe, should not be neg- 

 lected, and where the plan is available dried fruits may be kept. Plas- 



4S9 



