34 THE EXPERIMENTAL INDUCTION OF DEPENDENT NUTRITION. 



CISSUS-CARNEGIEA. 



Some freshly-cut slips of Cissus were inserted in cavities, some of which 

 opened on the crests of the ridges and some in the hollows between in 

 March, 1908. 



April 6, 1908.— Uost of the sUps were flaccid, but one was seen in which 

 a long root had been formed. 



April 10, 1908.— A tuber of Cissus wsls placed in a hollow cavity in the 

 terminal end of a branch, and several freshly-cut slips were inserted in 

 the same plant on the same date. 



April 25, /PO,?.— The insertions made last were all dying. 



April 25, 1908. — A second insertion of the original lot put in place in 

 March was developing a long aerial root. 



May 19, 1908. — The tips of all aerial roots were directed into punctures 

 made in the epidermal tissue of the host, but no action simulating that 

 ensuing when the roots enter the soil was observed. 



July 6, 1908. — The rigors of the arid foresummer had dried out nearly 

 all of the slips, but one was found to be alive when examined on October 6. 



April 20, 1909. — Several preparations had been made, but all had shown 

 no action except one in which a large swollen cutting of Cissus with broad 

 leaves was set down in a cavity on the surface of a stump of Carnegiea 

 and sealed with plaster. This was now developing a shoot several centi- 

 meters in length, which continued its growth during the summer at a 

 fairly normal rate, being again examined on June 12 and July 30, 1909. 

 During the early part of the dry foresummer the upper part of the stiimp 

 of the Carnegiea was surrounded with fine wire netting and a cloth shade 

 was arranged to protect the shoot of Cissus from the direct action of the 

 sun's rays. The swollen stem which had been placed in the hollow of a 

 branch on another plant in 1908 was now showing vigorous development 

 in several branches. 



October 5, 1909. — The Cissus on the stump of sahuaro had lost its leaves 

 and new branches, while the ones in the hollow end of a decaying branch 

 were showing luxuriant growth, with long, leafy stems. 



October 9, 1909. — The Cissus on the stump of sahuaro was developing a 

 bud at the base of the stem or branch which had died a few days before. 



November 9, 1909. — A dissection was made of the above arrangement, 

 and the Cissus was found to have made a large number of branching roots, 

 which ramified between the plaster and the cork formed by the host, not 

 effecting penetration in' any instance. The hard yellowish cork would not 

 permit the transfusion of more than the smallest amount of material, being 

 practically water-proof. 



It is to be seen that the stems of Cissus used as parasitic insertions 

 had a higher acid equivalent than the host on which they were placed, 

 while the salts dissolved in the sap were in greater proportion in the host 

 during the dry season. This condition is coupled with the fact that very 

 few of the arrangements made between these two plants showed anything 

 like a successful or enduring parasitism. 



The sahuaro accumulates a water-balance during the rainy season of the 

 summer that dilutes its sap to a concentration of salts below that of the 

 parasitic Cissus. While it is in this condition the parasite obtains some- 



