§2] UPON THE DIRECTION OF GKOWTH 337 



upwards towards the surface of tlie water — in response to the 

 more abundant oxygen supply there (Fig. 91), and will grow 

 along the surface of the water. Molisch undertook a sys- 

 tematic investigation of the action of various gases in con- 

 trolling this growth. 



The method employed was as follows : The gases were enclosed in glass 

 vessels whose mouth was closed by a plate of hard rubber perforated by 

 slits, 2 cm. long by 2 mm. broad. The vessel being laid on its side so that 

 the slits were vei-tical, the rootlet of a germinating grain was placed in front 

 of it. As the gas diffused from the vessel, it was for some time in excess 

 upon one side of the rootlet. The gases experimented with were pure 

 oxygen, pyrogallic acid, nitrogen, carbon dioxide, chlorine, hydrochloric 

 acid gas, illuminating gas, ammonia, nitrous oxide, ether, chloroform, and 

 oil of turpentine. 



In all cases there occurred, generally after about an hour, a 

 turning of the root towards the gas (positive aerotropism, 

 Molisch), followed by a marked curvature from tlie slit (neg- 

 ative aerotropism). Since decapitated roots respond in the 

 same way as intact ones, but in less degree, Molisch con- 

 cluded that the gases affect the growing region directly and do 

 not require the intervention of the root-tip. 



3. Chemotropism of Pollen-tubes. — The suggestion was early 

 made by Pfeffee ('83), as a consequence of his discovery of 

 chemotaxis in swarm-spores, that perhaps the bending of the 

 antheridium-tube of Saprolegnia towards the oogonium was a 

 case of response to a chemical agent. Steasbuegee ('86) 

 offered a similar suggestion for phanerogams. Pfeffee ('88) 

 then made experiments, but was unable to control the direction 

 of growth of pollen-tubes. Molisch ('89 and '93) was next 

 led to undertake further study in this direction by the obser- 

 vation that, when various pollen-grains are germinating in a 

 nutritive drop and a cover-glass is placed over them, the pollen- 

 tubes, after approaching near to the edge of the cover-glass, 

 turn away towards the centre again (Fig. 92, a, 6). The move- 

 ment from the margin of the cover-glass cannot be ascribed to 

 a difference of density produced by evaporation at the margin, 

 for it occurred in a saturated atmosphere ; nor can it be due to 

 surface tension of the bounding film of water, for the turning 

 occurred before the surface film was reached. These results 



