Apr. 1004.] BOTANICAL SOCIETY OF EDINBURGH 455 



considered a negligible quantity so far as tlie area of origin 

 of the branch rootlet is concerned. 



The ring on which the Lenticels may arise is then flat, 

 and the radially directed pressure must be practically the 

 same over the whole of its surface. It is precisely under 

 these conditions that Lenticels tend to occur above and 

 below the young rootlet, as well as laterally and obliquely. 



When, on the other hand, the branch rootlet is so large 

 relatively to the parent as to cause the ring-like surface of 

 contact to become an ellipse curved round the cylindrical 

 surface of the old root, then the conditions are considerably 

 altered, and the areas of minimal pressure come to be 

 situated at the sides of the lateral rootlet not above and 

 below it. 



In such cases lateral Lenticels alone are formed. 



That this rule is not constantly adhered to is probably 

 due to the abortion of some of the possible Lenticels, an 

 occurrence which may take place at any time, so that, 

 even when the rootlet is relatively quite thin, it often 

 happens that only two lateral Lenticels can be discovered, 

 or indeed only one may appear, either to right or left of 

 the base, or there may be an entire absence of Lenticels in 

 a moderately old root. 



The roots of a considerable number of plants were 

 examined with respect to the point of origin of their 

 Lenticels, but by far the most instructive preparations 

 were obtained from roots of Alder growing naturally in 

 water. In these the cortex remains in position for a much 

 longer period than in the case of roots from a drier environ- 

 ment, thus rendering the relations between the Lenticels 

 and the cortex of the lateral rootlet more easily discernable. 



That the exceptional mode of life of these roots does not 

 cause any abnormality in the Lenticels, so far at least as 

 their position is concerned, is shown by the occurrence of 

 similar relationships in roots from ordinary soils. It 

 is, however, considerably more difhcult to obtain good 

 preparations of these, as the cortex of the lateral rootlet is 

 usually cut off by the pericyclic cork very soon after the 

 primary division of the lenticellar initials has taken place. 

 The cortical tissue then dies, and at the same time shrinks 

 considerably, causing the Lenticels to lie apparently outside 



