72 CARNEGIE INSTITUTION OF WASHINGTON. 



is not a necessary prerequisite of successful parasitism in the case of 

 a species living under natural conditions. 



That the absorbing organs of a plant might withdraw liquids from 

 tissues of another plant, the sap of which had a higher concentration, 

 is also to be concluded from the results of recent work on imbibitional 

 phenomena at the Desert Laboratory. 



Extended series of measurements estabUshed the fact that a mixture 

 consisting of 90 per cent or more of agar and 10 per cent or less of pro- 

 tein, albumen, gelatine, tyrosin, or cystin takes up water in a manner 

 remarkably parallel to that of pieces of tissue of Uving plants. This 

 similarity is regarded as more than a coincidence. The plant proto- 

 plast consists largely of carbohydrates of the pentosan group, wdth 

 which are mixed varying proportions of nitrogenous material, which 

 may be in the form of protein, amino-acids, etc. Such a mixture 

 would have identical relations to water, either as swelling plates in the 

 laboratory or as sheets or strands of colloid in the cell. 



A number of agencies or conditions are found to affect the total 

 amount of water which may be taken up by this biocolloid mixture. 

 For example, some mixtures absorb slightly more water in acidified 

 solutions than in alkaUne, and many times as much from neutralized 

 as from either acid or alkaline solutions. Some salts in the solution 

 increase imbibition and some lessen it. 



These generalizations rest upon measurements made by the follow- 

 ing method: Small sections of dried plates of a mixture of biocolloids 

 were placed in trios in glass dishes into which various solutions might 

 be poured. Triangular pieces of thin glass were laid on these pieces. 

 The swinging vertical arm of an auxograph rested in a socket in the 

 middle of this plate. When the entire preparation was in readiness 

 and the pen at the other end of the compound lever was marking prop- 

 erly on the ruled paper of a revolving cylinder, the solution was poured 

 into the dish. The rate, course, and amount of expansion was recorded 

 by an inked line. (See Mem. N. Y. Botan. Garden, vol. 6, pp. 5-26, 

 1916, for a description of instrument.) 



The invaded tracts of the host are usually composed of expanded 

 vacuolated cells in which osmosis resulting from the solutions in the 

 vacuoles is the dominant hydrostatic agent, although the colloids sus- 

 pended in these vacuoles and the denser colloids of the cytoplasm have 

 their own imbibitional capacities. 



The younger cells of the haustorium which push into such masses 

 are probably not yet vacuolated. Absorption by them is almost en- 

 tirely by imbibition, and this would be carried on against any probable 

 osmotic action of a vacuolated cell. Thus, a thin plate of biocolloid 

 absorbed water from a solution of potassium nitrate which had an 

 osmotic coefficient of 60 atmospheres swelled about 400 per cent in 

 volume in 15 hours. 



