III.] THE CLASSIFICATION OF TIMBERS. 49 



ing between the medullary rays from centre to 

 circumference of the stem. In other cases sinuous 

 bands of small pores are seen running peripherally, 

 and almost simulating false rings, eg, the elm. 

 Beautifully-arranged tongue-like or flame-like groups 

 of pores are seen in the oaks, chestnut, Rhammis, 

 Ulex, &c., and these are very characteristic. 



Attempts have been made to carry the examination 

 of these features further by means of the microscope, 

 and to distinguish woods where the pores are single 

 from those where they are apt to be grouped in pairs, 

 threes and fours, and so on ; but although it is true 

 that the vessels are usually single in the box, for 

 instance, and often in groups of five to twelve or 

 more in holly and hazel, while less than twenty to 

 fifty together rarely occur in Rhamnus catkartzcus, &c., 

 I cannot find that the characters are either sufficiently 

 constant or sufficiently obvious for practical purposes. 



These, then, are the principal general characters 

 which can be employed in classifying timbers, and we 

 may now ask whether any others exist that could be 

 made use of. The reply is that several others could be 

 used more than they are if we had good records and 

 scales of comparison. Some of these may be shortly 

 indicated as follows : The hardness of different tim- 

 bers may be very different Thus the Indian Bonibax 



E 



