40 ROLAND M. HAEPER 



One who studies vegetation in this way does not need to know 

 the names of the plants, nor to worry about the pernicious ac- 

 tivities of species-sphtters and name-jugglers, which have filled 

 hundreds of pages of botanical magazines (but have been con- 

 sistently excluded from The Plant World through all its changes 

 of form and management). Of course there are several sources 

 of error to be guarded against, such as personal inadvertences, 

 errors of sampling, and variations of growth in different years 

 according to the weather, but some of these tend to counter- 

 balance each other, and others can be reduced to a minimum 

 by making numerous determinations in different spots and in 

 different years and averaging them. 



During the season of 1916 the writer made measurements of 

 a few easily accessible types of herbaceous vegetation in the 

 western part of Long Island, by way of experiment. In most 

 cases only one sample was taken, and that may not exactly 

 represent average conditions, but of course one does not usually 

 attain perfection in the first attempt. One square yard was 

 the unit of area selected, but the results are expressed in pounds 

 per acre, in order to facilitate comparson with smiilar studies 

 already published. (Readers who prefer metric units can 

 easily convert them.) 



Each square yard of vegetation was weighed as soon as 

 possible after cutting, and then allowed to dry for a month or 

 more, with frequent exposures to the sun (and with the assis- 

 tance of some artificial heat for some of the later ones), and 

 weighed again. A state of absolute dryness of course would 

 be better, but is not so easily attained. Determining the ash 

 involved some minor difficulties, for the organic matter in hay 

 will not burn up completely of itself, but requires the applica- 

 tion of some external heat; and suitable incinerating apparatus 

 is not to be found in most homes or even in botanical labora- 

 tories. Small amounts of ash were probably lost in the impro- 

 vised apparatus used, but that does not materially affect the 

 comparison between different samples. 



The vegetation types studied are as follows: 



