DEPARTMENT OF BOTANICAL RESEARCH. 73 



One of these designs was taken for immediate development. Six 

 instruments were completed early in the year, and these were attached 

 to trees at Baltimore, St. Louis, Tucson, Santa Catalina Mountains, 

 Arizona, and Carmel, California, as described below, with the result 

 that the seasonal activities of six species of trees during one season are 

 now included in the records. 



These measurements were made of the bases of trunks of trees 30 to 

 45 cm. in diameter at a height of 1.2 to 1.5 m. above the ground, and 

 are therefore directly comparable to the figures obtained by the tape- 

 line and calipers of the forester. 



The variations of the base of a trunk are, however, but an imperfect 

 index of the activities of a tree, and any adequate study of the physics 

 of growth, the formation of wood, the development of bark, or the 

 ascent of sap must be based upon measurement of the entire trunk or 

 "cone" of the stem. It has been important, therefore, to construct 

 an instrument which might be attached to trees at various heights up to 

 50 or 60 feet above the base, which would be so perfectly automatic 

 as to need a minimum of attention. This has been accompHshed, as 

 evidenced by the fact that an instrument attached to a Chihuahua pine 

 at an elevation of 6,000 feet, at a distance of 22 miles from the Desert 

 Laboratory, was put in place during the first week in April and has 

 made a perfect record during the season of 200 days, with no attention 

 except a weekly visit to remove and replace record shps and wind the 

 clock of the recording drum. 



The essential part of the dendrograph is a "yoke" constructed of 

 strips of some alloy, such as bario or invar, with a low temperature 

 coefficient, 25 by 0.2 cm., arranged in the form of a U or of a poly- 

 gon in such manner as to inclose the tree-trunk, with a clearance of 

 2 to 5 cm. on all sides. At one side (on the curve of the U) two contact- 

 screws work in a horizontal position to a suitable pressure on seatings 

 in the bark. The bearing end of a short lever seated on the instrument 

 bar which bridges the end of the U or closes the polygon makes a con- 

 tact on the opposite side, and any variations in the diameter of the 

 tree between the end of the lever and of the two contact-screws (con- 

 sidered as one) will cause movements of the lever which will be trans- 

 mitted to the pen arm which traces an inked hne on the recording drum. 



The yoke and the recording instrument are supported by a belt of 

 hinged wooden blocks which have been boiled in paraffine. Sections 

 of spring brass 1 mm. in diameter and 8 cm. long, fastened in small 

 brass plates seated on the blocks, form a flexible support for the yoke 

 and hold it with an adjustable tension. The recording instrument, of 

 a design recently perfected for this use and with the auxograph, is 

 seated on a large block which closes the circle of the belt. The lever 

 system of the recorder engages the free end of the bearing lever on 

 the instrument bar of the yoke and records all variations in diame- 



