204 



SCIENCE 



[N. S. Vol. XXXVIII. No. 971 



under and upper sides of branches. These two 

 investigators agree as to the main groups of 

 trees in respect to the effect changes of tem- 

 perature have on the position of their 

 branches. It seems, therefore, that Ganong 

 happened to use trees and shrubs which be- 

 longed to only one of these classes. The expla- 

 nation advanced by Caspary is suggestive be- 

 cause it is based on a differential longitudinal 

 contraction of the wood in branches. Some 

 of his earlier studies" have shown that tree 

 trunks undergo transverse contraction in pro- 

 portion to the degree of cold and that the 

 assumptions to the contrary are incorrect. 

 That longitudinal contraction of wood takes 

 place as the temperature is lowered is upheld 

 by many general observations. Trees are fre- 

 quently cleft in forks of the trunk during 

 winter and these clefts open when it gets cold 

 and close as wanner weather comes. In an- 

 other connection the writer found that 

 crotch clefts were always at right angles to the 

 branching and usually widest above, appearing as 

 though the crotches had been split by driving in 

 a thin wedge from above. In two instances where 

 measurements were taken the component parts of 

 the crotches had separated about 2 em., which 

 seems to indicate that there had also been a longi- 

 tudinal contraction of the outer portions of the 

 trunks, thus resulting in an outward bending of 

 the branches." 



Caspary's observations on the lateral dis- 

 placement of some tree branches also fit into 

 his contraction theory, although he failed to 

 note it, provided it is assumed that the trees 

 on which this movement occurred had trunks 

 with the so-called twisted grain, for in such a 

 case longitudinal contraction would necessarily 

 result in lateral movement of the attached 

 branches. 



In this connection it seems of interest to 

 notice some of the peculiarities of arrange- 

 ment of the tissues about the bases of branches 

 that were studied by Jost." He found that 

 the cambium at the basal angles of branches 



' Loc. cit. 



"" Loc. cit., pp. 36-37. 



" L. Jost, ' ' TJeber einige Eigenthttmlichkeiten 

 des Cambiums der Baume, " Bot. Zeit., 59: 1-24, 

 1901. 



is not eliminated as the stems and branches 

 grow in diameter, but that its cells and those 

 of the tissues differentiating from the cambium 

 glide between each other and also become 

 shorter. In case of the adaxile side crowding 

 and compression are more marked than on the 

 abaxile side, apparently because the angle is 

 usually much smaller. Sometimes the bark in 

 the adasile angle is not forced outward, but is 

 included, and under such conditions the pres- 

 sure in the angle compels the cambium under 

 the included bark to cease growth. Most 

 commonly, however, the wood-growth in the 

 angle forces the bark outward and thereby 

 induces a more rapid reduction in the cambial 

 area and a greater increase in thickness per 

 annual ring than on the abaxile side. In 

 addition to gliding between each other, the 

 cells in the adaxile side are turned at a 

 tangential angle so that large groups of them 

 come to lie almost horizontal or at right angles 

 to the axis, while groups of cells from the 

 branch and from the stem sides are forced in 

 among these transverse cells of the crotch. 

 Usually, then, no cambial cells are eliminated 

 in branch-angles, but they are forced between 

 their neighbors and complicated tangles re- 

 sult in which often large groups of cells come 

 to lie in a more or less transverse direction. 

 The ends of medullary rays vertical to each 

 other in the base of branches come closer to- 

 gether and may even cross each other. 



In view of the fact that the groups of par- 

 tially transverse tissues at the base of a branch 

 are probably under more or less pressure and 

 because changes of temperature have a much 

 greater effect upon transverse than upon 

 longitudinal dimensions it seems possible that 

 the differential contraction which according 

 to Caspary is the cause of the thermometric 

 branch-movements may be chiefly confined to 

 the bases of branches and depend upon these 

 peculiar gnarly growths described by Jost, 

 and perhaps their arrangement about the base 

 of a branch which is usually characteristic 

 for a species, may determine whether a branch 

 shall move up or down as the temperature 

 sinks. The relative amounts of " spring " and 

 " summer " wood in the under and upper sides 



