SECTIONAL TRANSACTIONS.— K*. 395 



time and money on the production of willow timber, many sound trees 

 were rejected by buyers, as unsuitable for bats. 



Preliminary investigations showed that soil, soil-moisture, or climate 

 could not be the cause of the trouble, since trees growing on the same soil, 

 and alternately with trees accepted by buyers, were rejected, although 

 perfectly healthy, sound and straight, and of the right size. 



There were indications, however, that the differences in quality might 

 be correlated with botanical differences. It was found that at least four 

 different strains of Salix alba var. ccerulea occur in plantations of cricket- 

 bat willows. 



Methods of cultivation are described. Attention is called to the need 

 for further systematic and for genetical work, and for thorough practical 

 tests of the timbers of the different strains, of both sexes. 



The paper is illustrated by lantern-slides showing the habit-characters of 

 the tree, the conditions of cultivation, and the catkin peculiarities of the 

 several strains. 



Afternoon. 

 Excursion to Ribston Hall Estate. 



Saturday, September 3. 



Excursion to Rievaulx and Gilling. 



Sunday, September 4. 



Excursion to the Forestry Commission, Allerston. 



Monday, September 5. 



(Meeting with Section K, Botany, q.v.) 



Tuesday, September 6. 



Miss M. M. Chattaway. — Tile-cells in the rays of the Malvales. 



Certain genera of the Malvales have rays containing a peculiar type of 

 cell (' Ziegelsteinformige,' Moll and Janssonius) which has not been observed 

 elsewhere. These cells are apparently the result of more frequent tangential 

 divisions than the normal ray cell, and differ from the erect cells of hetero- 

 geneous rays in shape and position in the ray. They are devoid of contents 

 and can be divided into two types. Type A, tile-cells two or more times as 

 high as the procumbent cells ; usually not more than four cells correspond 

 to one procumbent cell (radial section) ; conspicuous on radial and tangential 

 sections, but not on transverse. Type B, tile-cells equal in height to the 

 procumbent cells ; often eight to twelve cells corresponding to one pro- 

 cumbent cell (radial section) ; conspicuous on radial and transverse sections, 

 but not on tangential. 



An attempt has been made to follow the development of these cells from 

 the cambium in fresh material, and to discover their function in the rays of 

 the Malvales. Comparison with genera from other families in which rays 

 without procumbent cells occur shows that their cells differ in structure 

 from tile-cells, and probably also in development and function. 



