July 13, 1SS3.] 



SCIENCE. 



39 



only in all its proportions the noblest of all 

 our coniferous trees cast of the Kocky Moun- 

 tains, vying in girtli and height with the yellow 

 poplar (the Liriodendron tulipifera of the 

 botanists), but it is by far the most stately 

 all tlie of trees belonging on tlic eastern face of 

 the continent. Moreover, it has certain haliits 

 which are altogether peculiar to its species, 

 anil which constitute it a very exceptional 

 member of the C'oniferae. 



When this tree grows on the dry ground, 

 or on a surface where the water does not stand 

 during the summer half of tlic year, it ditlcrs 

 in no important feature from its kindred 

 species ; but, when it grows in swamps which 

 are flooded during the spring or summer 

 months, the roots form excrescences, which rise 

 so that their crests overtop the level of the 

 water during these seasons. These excres- 

 cences are of varying height, their projection 

 above the level of the roots depending on the 

 depth of the swamp-waters during tliose sea- 

 sons of growth. These conditions may be 

 satisfied by projections, or ' linees ' as thoj' 

 are called, that rise only a few inches above 

 the root, or they may rise to the height of five 

 or six feet above the soil. These knees are 

 sub-cylindrical in form ; near the base they are 

 elongated in the direction in which the root 

 extends ; above, they give a nearlj' circular 

 section ; at the top thev are crowned by a 

 cabbage-shaped expansion of bark of irregular 

 shape, rough and warty without, often hollow 

 ■within. They are often as much as eighteen 

 inches in diameter. Tiic}' are so commonly 

 hollow, and of such size, that they are sometimes 

 used by the natives for beehives or for well- 

 buckets, for either of which uses they are toler- 

 ably well adapted. A tree of large .size, say 

 sis feet in diameter, will often have as many 

 as thirty or fortj- of these knees projecting 

 above the swamp-water which surrounds its base. 



Looking closely at these knees, we observe, 

 that, unless they are evidently decayed, tliey 

 generally have a verj- porous, spongy bark 

 over the surface of their crests ; and the bark 

 on this summit, peeling off from time to time, 

 often exposes a singularly spongy surface, such 

 as we find in the inner bark of the pine-tree 

 when tlie coarse outer bark is peeled away. 



There have been many conjectures as to the 

 function of these knees. It has been supposed 

 that thej- were in the nature of suckers or 

 branches from the roots, which gave rise to 

 new trees ; but, after examining thousands of 

 these knees, I am convinced that they never 

 have this nature. In no case have I seen or 

 heard of any buds appearing on them. The 



only clew to their function I liave obtained 

 in the following way : whenever it happens that 

 tlie knees become entirely submerged during 

 the growing season, the trees to which tiiej' 

 belong inevitably die. Very extensive proof of 

 this point was given by the general submer- 

 gence of extensive districts during the earth- 

 quakes of 1811-13, in tlie region near the 

 Mississippi, wliere the cypress-trees over a 

 region several hundred miles in area were 

 killed by a subsidence that brought the water 

 a foot or two below the crests of the knees. 

 In Keel-Foot Lake, in Kentucky and Tennes- 

 see, thousands of these long ordinary cypress- 

 boles still stand in the shallow waters, though 

 it is now seventy- years since they were killed 

 by the sligiit submergence of their knees. 



The same thing can be seen on a smaller 

 scale in several mill-ponds in western Ken- 

 tucky, where the change in level of the swamp- 

 water has brought these excrescences below 

 the surface of the water. These facts — viz., 

 the absence of the knees when the tree grows 

 on high land, and the death of the tree when 

 the knees are permanently submerged — lead 

 me to the opinion that the use of these ex- 

 crescences is to bring the sap while in the 

 roots in contact with the air. That they have 

 this function is made more probable b}' the 

 fact that their heads, i.e., the parts which 

 always project above the water during the 

 growing season, remain very vascular, and, bj' 

 a process of desijuaraation, secure the expos- 

 ure of the inner bark to the air. 



It is evident that this tree atfords us a very 

 interesting instance of a specialized structure, 

 that only develops when the plant occupies a 

 certain position. We often find tiiis tree arti- 

 ficially transplanted to the gardens of the 

 western country. It tlien shows no distinct 

 tendency to form knees, though the surface of 

 the roots show a few short spurs not over an 

 inch or so high. 



It is a well-known fact that the genus Tax- 

 odium dates back into the early tertiaries. I 

 am not aware, however, that fossil knees have 

 ever been found. AVe have only to examine 

 the borders of the swamps to see that it can- 

 not, on the uplands, maintain a battle with the 

 contending broad-leaved trees, though in anj- 

 artificial open place it will grow with singular 

 luxuriance. 



It seems to me likely that wc have here a 

 very interesting case of a species owing its 

 survival to a peculiar habit of growth. There 

 can hardly he a doubt that the kindred of this 

 Taxodium held an important place on the 

 continent before the development of the broad- 



