Dec. 1891.] THE BOTANICAL SOCIETY OF EDINBURGH. 225 



The depth to which the roots penetrate in search of food 

 is fairly well shown by the two columns indicating the 

 amounts of root matter above and below the 8 -inch limit. 



Pre-eminent among deep-rooting grasses is rye-grass, which 

 has more than half of its roots below the 8 -inch layer. 

 This is one of the most valuable characteristics of that grass. 

 In the short period of one year its roots are able to pene- 

 trate to the subsoil, and bring up to the surface the stores 

 of plant food contained there, some of which would otherwise 

 be lost to the soil. 



That this is the habit of the species is seen by the fact 

 that the two samples of rye-grass resemble each other in 

 that respect, although differing so widely in quality. 



Next in value in this respect is the meadow fescue, 

 nearly the half of whose roots dip below 8 inches, and then 

 the tall fescue, fully one-third of whose roots are in the 

 lower soil and subsoil. Even the common meadow-grass 

 {Poa pratensis) has more than one-fourth of its roots in the 

 lower soil, and in that respect it is a serviceable grass, but 

 the other poas have a very feeble hold upon the subsoil, and 

 belong to that class of grasses which are easily injured by 

 extremes of heat and cold. 



As to the sweet vernal grass (Anthoxantimm odoratum), 

 it is seen to be a merely surface grass, finding with diffi- 

 culty its nourishment in that part of the soil where it is 

 subjected to the greatest competition from its stronger 

 rivals. Fortunately it is not a grass that it is desirable to 

 have in a pasture except to a very limited extent — more as 

 a flavouring ingredient than as any considerable component 

 of the grass or hay. 



The specimens of the actual roots that are before you tell 

 their story more eloquently than the diagrams, but not so 

 accurately, as the eye cannot do more than examine their 

 surface. The roots are gummed on to strips of paper exactly 

 the width of the box in which they were grown. You will 

 see that they are there in their entirety, and as nearly in situ 

 as it is possible for that which is grown in three dimensional 

 space to be when it is flattened down into a space of two 

 dimensions. 



In conclusion, I would again remind you that these 

 observations are made upon grass of one year old. A set of 



