FUNCTION OF SIEVE-TUBES. COMPANION-CELLS 335 



siderable number of sieve-plates. Thence we may infer that any 

 differences of pressure which arise within the intact sieve-tube system, 

 owing to the partial depletion of the tubes at certain points, are at once 

 equalised by a more or less rapid displacement of the liquid contents in 

 the corresponding direction. 



Whether the hydrostatic pressure in the sieve-tubes owes its origin 

 to the osmotic properties of the liquid contents, or whether it is due to 

 compression of the sieve-tubes by the highly turgescent adjoining tissues 

 (leptome-parenchyma and companion cells), is still uncertain. Most prob- 

 ably both factors have a share in producing the pressure observed. If 

 sieve-tubes are able to develop a turgor-pressure on their own account, 

 this capacity is entirely due to the presence of a peripheral layer of 

 cytoplasm (or at any rate of a plasmatic membrane). It is, indeed, 

 conceivable that the living contents of sieve-tubes are mainly concerned 

 with the development and regulation of pressure. Strasburger, on the 

 other hand, believes that the peripheral protoplasm serves to retain the 

 substances which are undergoing translocation within the confines of 

 the sieve-tube, and thus to prevent them from diffusing into the 

 adjoining cells. Czapek finally supposes that the living protoplasm 

 plays some active part in the process of translocation quite apart 

 from its role as the producer of turgor-pressure. This matter evidently 

 requires further investigation. 



Translocation of the starch which is often contained in sieve-tubes 

 has never been observed ; it undoubtedly only takes place in very 

 exceptional circumstances. According to Strasburger, the presence of 

 starch is in some way connected with the formation of callus, growth of 

 the callus-pads being correlated with disappearance of the starch. 

 Besides proteins and starch, sieve-tubes often contain various sugars, 

 and therefore probably take some part in the transportation of carbo- 

 hydrates ; the bulk of the carbohydrate material, however, certainly 

 travels in the conducting parenchyma. 



Among Angiosperms the primary mother-cells of sieve-tube segments 

 as a rule undergo several divisions. This process was first described by 

 De Bary, and was afterwards studied in greater detail by Wilhelm and 

 others. The largest of the resulting daughter-cells becomes the actual 

 sieve-tube segment, while the rest develop into the much narrower 

 companion-cells, as these sister elements are termed by Wilhelm. (In 

 the bundle-ends the relative sizes of sieve-tubes and companion-cells 

 are reversed.) A companion-cell (Fig. 136, g) may be readily dis- 

 tinguished from ordinary leptome-parenchyma (cambiform cells) by its 

 narrow cavity, by its abundant protoplasm and large nucleus, and more 

 particularly by the fact that the wall separating a companion-cell from a 

 sieve-tube always bears numerous transversely elongated pits which 



