Mariotte, so well known by his researches into the compressibility of air, represented the rise of the sap as dependent upon what he called "the attraction taking place in the narrow tubes"--upon what, in fact, we now term _capillarity_. "This first entrance of water into the roots is in obedience," he said, "to a law of nature; for wherever very narrow tubes exist which touch the water, it enters into them, and even rises, contrary to its natural inclination." * * * * * Many botanists adopted the opinion of Mariotte. But if it were well founded, all _capillary_ bodies, even inorganic ones, ought to present a circulatory movement analogous to that of the sap. Now, this is not the case. A body must be animated, must be living, for attraction to take place in the narrow tubes, and to produce a movement comparable to that of the nutritive liquid. * * * * * Malpighi attributed the rise of the sap to the alternating rarefaction and condensation of this liquid by heat; Perrault, to a kind of fermentation; De Saussure, to a peculiar irritability of the vessels. Of these three hypotheses, the first is purely physical; the second, chemical; the third, vital. So, as we see, there is something for everybody--_chacun à son gout!_ * * * * * The same question has, in our own day, been taken up from a new point of view, on the occasion of Dutrochet's discovery of the endosmose. This philosopher was one of the first to perceive that two liquids, separated from one another by a membrane, quickly effect or induce a current which always carries the thinner liquid towards the denser, and ends by mingling the two completely. "It is endosmose," he said, "which produces at one and the same time the progression of the sap by _impulsion_, and its progression by _affluxion_. The sap would receive its impulse in the spongioles of the roots; thence would be carried towards the upper parts by the turgescence of the organs--by the affluxion, which would thus act as a forcible mode of suction." The basis of this theory is, that the sap contained in the upper parts will be more concentrated or denser than that in the lower portions of the same plant. But this is a mere supposition. And even this supposition has been swept away by the recent experiments of Hartig and others, which show that the difference in density between the two saps is not only almost null, but in many ligneous plants the lower sap is, on the contrary, denser than the upper.[41] * * * * *