48 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



adjacent radial columns ; for instance, at one place 1940 is made up 

 of one row of wide plus one row of narrow cells, at another place 

 three rows of narrow cells constitute the densewood of 1939 plus all 

 of 1940. The growth layer for 1940 has other weaknesses too; these 

 are below normal lignification and wall thickening. When we move 

 outward and consider 1941, we find its lightwood to be large celled 

 and thin walled, thus making a decided contrast with all of 1940 as 

 well as the densewood of 1939. If these growth layers were not dated 

 by absolute methods, many workers would hesitate to call 1940 an an- 

 nual increment. Moreover, a lack of close study, not only of this 

 specimen but also of the many others we have had, would merit a con- 

 clusion that the cambium had experienced a period of dormancy ex- 

 ceeding a year, whereas close study indicates that the cambium was 

 sluggish, probably very sluggish, but not dormant. It is worthy of 

 note that on tree TTP 23 the cambium in some limbs was highly 

 active, in others like TTP 23-4 it was rather sluggish. This applies 

 with equal emphasis to different areas longitudinally and tangentially 

 in the same branch. 



Marginal definition is impaired, finally, by the presence of a transi- 

 tion from densewood outward into lightwood. In nearly all instances, 

 the indefinite contact is due either to subnormal or to abnormal de- 

 velopment of the initial lightwood laid down. This type of feature as 

 a cause of diffuseness is more elusive, more confusing, and more con- 

 ducive to errors of identification, probably, than any of those already 

 discussed. The true features of densewood and lightwood are suf- 

 ficiently masked to give the student a feeling of uncertainty, and yet 

 they are present with a definiteness to challenge his eyes and his in- 

 terpretative ability. Back of it all stands the problem of physiological 

 processes as recorded by cambial activity. 



Among the angiosperms, growth-layer contacts are at best far from 

 being as clear cut as they are in gymnosperms. Even so, there are 

 varying degrees of diffuseness. The Siberian elm gives a good ex- 

 ample. In four different branches of SE i the contacts are very in- 

 definite and difficult to locate with precision. They can be located with 

 the use of high power over most of the circuit, but over an arc of more 

 than 90° they are completely indistinguishable. One would be tempted 

 to interpret the growth layers as "false," that is, diffuse intra-annuals, 

 were it not for a late frost injury which occurred in 1938. The con- 

 tact immediately inward from the frost, annual of necessity, possesses 

 a diffuseness rivaling that of the others. Another point : a hand lens 

 revealed two complete, entire growth layers whereas the microscope 

 showed five. Thus a hand lens can give rise to gross errors. 



