264 Eepokt of the Botanical Department of the 



other side of the cleft it had died back only about 4 inm. from the 

 cleft edge. 



The Bismarck tree referred to above had a radial cleft in the loose 

 bark on the northwest side 17 cm. long, and another 7 cm. long in a 

 loose patch opposite the large one. The edges of the larger cleft 

 stood apart 6 to 7 mm. in the middle, and the bark had died back 

 about 6 to 10 mm. on both sides of the crack. Figure A on Plate 

 VTI is a photograph of this tree trunk, taken May 7, 1911. Without 

 moving the camera all the loose bark was removed and the tree re- 

 photographed; but the negative was spoiled, and therefore another 

 was made the next day. As may be seen from figure B the camera 

 was a little closer, although an attempt was made to set the tripod 

 legs into the same holes they had occupied when figure A was made. 

 A number of very fine wind-checks are also noticeable on the exposed 

 surface. 



These figures show that the cleft did not extend to the ground, 

 although it went several centimeters below the surface of the conical 

 mound of soil which had been heaped about the tree during winter. 

 The lower end of the shorter cleft, on the opposite side of the trunk, 

 began at the same height as the lower end of this cleft. The longer 

 cleft extended practically the entire length of the loose patch of bark. 

 On May 7, the tree measured 23.5 cm. in circumference. After 

 removing all the loose bark on both sides of the trunk only two 

 narrow bands of adhering bark were left between them. That on 

 the northeast side measured 3.75 cm. and the one on the southwest 

 only 2 cm., thus leaving a quarter of the bark adhering to the wood 

 as bridges. 



The loose bark was alive except along the margin of the cleft, and 

 seemed to have increased slightly in thickness, for its iimer surface 

 was greyish white and glistened like a regenerating surface. The 

 exposed wood surface had a similar appearance except that an irregu- 

 lar, band-like region, exposed to the air by the crack in the bark, was 

 brown; and that a few other browoiish spots were scattered irregu- 

 larly over the other portions of it. But, even though tl is wood 

 surface was exposed to the air about 24 hours before being covered 

 with grafting wax, considerable regeneration occurred later, as shown 

 in figure A of the next plate, by the bursting of the wax cover (August 

 15). On September 11 the wax was carefully removed to determine 

 the amount and appearance of the regeneration which had occurred. 

 Figure C of Plate VHI shows that a narrow strip of wood surface ex- 

 posed by the cleft as well as various other places failed to regenerate 

 or did so only partially. At that time the tree measured 28 cm. in 

 circumference. The bridging band of bark on the southwest side 

 had increased to about 5.5 cm. in width by lateral growth, and that 

 on the northeast side to about 8 to 9 cm., although its lateral limits 

 were not easily made out on account of the almost perfect regen- 

 eration taking place on the adjoining surface of exposed wood. 



