286 



D. H. PAGE 



Fig. 3. Carbon replica, metal-shadowed, of a beaten spruce 

 tracheid and its associated fibrillation. Magnification 1500, 

 inset 13,000. 



Fig. 4. Carbon replica, metal-shadowed, of a surface of wood 

 tissue. The wood — Scots pine — was cut in a radial plane 

 with a chisel prior to replication. Magnification < 1700. 



remain embedded in the methacrylate ic). The 

 stripping of the fibres is sometimes dil'Hcult, cellulose 

 tape being by no means always effective. A more 

 successful method employs a 10 "„ solution of poly- 

 vinyl alcohol (PVA) in water. This solution is poured 

 onto the methacrylate cast and gradually sets to a 

 hard film which when stripped, brings with it the 

 embedded fibre id). In particularly stubborn cases 

 this process can be repeated. 



From the negative cast so formed a robust positive 

 replica is made in PVA. A 10 % solution of PVA in 

 water, sufficient to form a film about 0.5 mm thick, 

 is poured onto the cast and allowed to set (e). This 

 replica, which is a reproduction of the original fibre 

 lying on its glass substrate, can be separated easily 

 from the methacrylate (/). It is preshadowed with 

 gold, palladium, and carbon is deposited on it in 

 the usual manner. 



At this point a handling technique is adopted (5) 

 that ensures that the fragile carbon replica is mount- 

 ed intact on its grid. Briefly this technique consists 

 of embedding the carbon replica together with the 

 support grid in thick plastic films which are washed 

 away after the composite arrangement has been 



inserted in the specimen cap and holder. Thus the 

 delicate replica surface is at no time exposed during 

 the manipulation of the grid and accidental damage 

 to the replica is avoided. The only chance of damage 

 is during the final flow-wash (a very gentle one) with 

 chloroform. Very thin replicas of rough surfaces 

 can be made and mounted on coarse mesh grids 

 with confidence. For the replication of fibres, carbon 

 films 70 A thick (estimated by a method based on 

 optical density measurement due to Agar (1)) have 

 been used. "New 100" grids made by Smethurst 

 Highlight Ltd., which have apertures about 200 

 microns square and whose grid bars obscure only 

 about 13 "o of the total grid area, are suitable. 

 Under such conditions replicas have invariably been 

 intact over the whole area of the grid. 



Some of the results obtained are shown in the 

 micrographs of figs. 3 and 4. Fig. 3 shows at a low 

 magnification part of a spruce sulphite tracheid 

 that has been mechanically beaten. The ribbonlike 

 piece of fibrillation coming away from it is of interest 

 and the indicated region is shown inset at a higher 

 magnification. The compactness and uniformity of 

 direction of the microfibrils strongly suggest that 



