pear an exception to the general rule respecting buds and roots; but the author observes, that the tuber differs but little from a branch which has dilated instead of extending itself. The runners, which give existence to the tubers beneath the soil, are very similar in organization to the stem of the plant; and if exposed, readily emit leaves, and perform all the functions of the stem; and. on the other hand, Mr. Knight has shown, in a former memoir, that the buds on any part of the stern may be made to produce tubers similar to those formed beneath the soil; but he has never, under any circumstances, been able to obtain tubers from the fibrous roots of the plant.
Many naturalists have imagined the fibrous roots of all plants to be of annual duration only, because those of bulbous and tuberous plants certainly are so; but Mr. Knight observes, that the organization of trees is extremely different; and he has not found any portion of their roots to be deciduous.
The author, having observed a new species of joint in the Squalus maximum of Linnaeus, takes occasion to trace the successive gradations of a similar structure, through various kinds of fish, to the more remote resemblance to be found in quadrupeds and in men.
In the Squalus each joint of the spine approaches, in some measure, to that which is termed the ball and socket joint, as a concave surface of each vertebra is applied to a ball; but the ball, in this instance, is not, as in other cases, a smooth surface covering a solid bone, but a collection of fluid contained in a bag that is nearly spherical, round which the concave surfaces of the vertebrae are moved.
In a fish of thirty feet in length, the diameter of the body of one of the largest vertebrae measured seven inches; the quantity of fluid in one of the cavities amounted to three pints; the ligamentous substance, which unites the vertebrae, being nearly one inch in thickness, externally very compact and elastic. but internally possessed of but little elasticity.
The elasticity of these ligaments preserves the straitness of the spine when it is not acted upon by the muscles, or by other external force; and though the extent of motion, in any one joint, must be small, their number affords considerable latitude of motion.
Since the vertebrae, in other fish, are found with concavities in each surface, it was natural to expect a corresponding resemblance in the intervertebral structure; and in the skate this was found to be the case, and the cavity nearly spherical, as in the Squalus. In the common eel it is more oblong, the longitudinal diameter exceeding the transverse one by about one third.
In the sturgeon the structure varies considerably, as the cavities communicate with each other by apertures through the bodies of the vertebrae, which in this fish are cartilaginous rings, connected toge-
