• Scallan, A.M.
  Pulp and Paper Research Institute of Canada
  
• Green, H.V.
  Pulp and Paper Research Institute of Canada
  

For softwoods, fibre width can be calculated from a count by optical microscopy of the number of cells per unit area (N) in the cross-sectional face of a small block of wood. Assuming a square cross section for the average fibre, the fibre width (b) is given by b = (1/n)^1/2. Lumen width (a) can be calculated from the fibre width thus obtained amd a measurement of the bulk density of the wood (D_B) using the theoretically derived relationship a/b = (1-D_B/D_C)^1/2.D_C is the cell-wall density, which to a good approximation is a constant from one wood to another.

For hardwoods, the count of cells per unit area is restricted to areas occupied by the libriform fibres, and the bulk density used is that of the part of the wood occupied by these fibres. This bulk density may be calculated from the bulk density of the whole wood using a measurement of the fraction of the total volume occupied by vessels and ray cells.

The results so obtained are in close agreement with those obtained by direct measurement using scanning electron microscopy and are believed to superior to those previously obtained by the usual expedient of direct measurement by optical microscopy. This is particularly true of cell-wall thickness, (b-a)/2, the direct measurement of which has recently been shown to be subject to many sources of error.

A simple experimental procedure for measuring the dry-bulk density of small samples of wood is described and the work is illustrated by measurements of the fibre width, cell-wall thickness, and fibre coarseness of some thirty species of wood.