SHUFELDT] ACCURACY IN DESCRIBING TREES 293 



summit into coarse and straggling branches." This may be tlie 

 case in some tulip trees, but it by no means applies to all of them. 

 For instance, it in no way applies to the tree I here present in 

 Figure i, which is a reproduction of one of my photographs of a 

 superb Liriodendron tulipifera, growing on a hill south of Pierce's 

 Mill, Washington, D. C. It is on the right hand side of the road 

 before coming to the bridge close to the mill, and not far from the 

 boundary of the National Zoological Park property. It will at 

 once be observed that, although the trunk of this tree is fairly 

 "straight," it is by no means ''clear" of branches; neither does it 

 exactly ''divide rather abruptly at the summit." Personally, not 

 only have I seen many a tulip tree like this one, but I have also 

 seen specimens wherein the limbs, coming off from the trunk, were 

 much larger, and sprang from the trunk much lower down. More- 

 over, the trunk of this tree is not always "straight," though it is 

 so in the majority of trees of this species. (See picture on cover.) 



There is a very excellent description of this tree in the Centtiry 

 Dictionary (Knowlton?), but it likewise ignores the jorm of it. 

 This definition, however, gives a very interesting fact when it 

 states that "It is the sole remaining representative of a nearly 

 extinct type which was formerly abundant, not less than 17 fossil 

 species being known, the greater part occurring in the Cretaceous 

 formation in New Jersey, Kansas, Nebraska, Wyoming, Greenland, 

 and Bohemia, with a few in the Tertiary, chiefly in Europe." 



Large tulip trees are very conspicuous in our forests where they 

 occur, especially when they are in blossom, and their trunks are 

 large, clear of branches a long distance up from the ground, 

 markedly straight, and gradually tapering to their lowermost 

 branches. These facts should not be ignored in scientific descrip- 

 tions of Liriodendron tulipifera, much less should the exceptions to 

 these general characters be omitted. 



I find these loose descriptions and these omissions, in the case of 

 many of our American trees, all the way through our standard 

 botanies ; and when the trees are entirely unknown to the student 

 of such species, and the works give incorrect and only partial 

 descriptions, the latter fail utterly to help him as they should, and 

 science falls short of its purpose to just that extent. With respect 

 to the general characters of Liriodendron tulipifera I may say, that 

 the flowers show three reflexed sepals and that there are six petals 

 (Fig. 2), these latter being arranged in two (2) rows, thus forming 



