140 ECONOMIC WOODS OF THE UNITED STATES 



pits may be very small, medium or large, often with considerable 

 variation in the same specimen. In woods with heterogeneoas 

 rays the marginal cells are usually more prominently pitted than 

 the others. In Gordonia and Oxydendrum the pits are simple 

 or only sUghtly bordered and are frequently in scalariform ar- 

 rangement. In Sideroxylon and Chrysophyllum many of the pits 

 are small and bordered while others arc large, simple or nearly so, 

 elliptical or elongated-elliptical and disposed horizontally, verti- 

 cally or diagonally, resembling perforations rather than pits. In 

 Magnolia it is common to find much elongated borders about 

 groups of small pits. 



Table X gives for the different families the dominant type 

 of pits in vessels where in contact with the rays. Where both 

 types are indicated with connecting line it refers to their occur- 

 rence side by side in the same wood ; otherwise in different woods 

 of same family. An arrow indicates transitions from the pre- 

 vailing type. 



VESSEL CONTENTS 



The principal contents of vessels that have ceased to function 

 actively as water-carriers are (a) tyloses (parenchymatous in- 

 trusions) and (6) various deposits or excretions such as gums, 

 resins, lime, etc. Sometimes such features are constant and con- 

 spicuous enough to be of value for diagnostic purposes. In a great 

 many cases, however, there is too much variation for dependable 

 results. Generally it is merely a question as to whether the pores 

 appear open or closed rather than exact determination by micro- 

 scopic means of the presence or absence of certain contents. The 

 feature is of most importance in woods with large pores. 



The following table gives the results of some investigations by 

 the author on the occurrence of tyloses and gum deposits in in- 

 digenous woods and a few that have been introduced. The find- 

 ings do not in all cases agree with those of other investigators and 

 in some instances are not to be considered as final, especially where 

 non-occurrence is indicated, owing to the great likelihood of varia- 

 tion in different specimens. 



