168 TRANSLOCATION IN PLANTS 



gradient of 7.0 atmospheres (0.14 atmosphere per centi- 

 meter) would be 3.3ju. If the flow were taking place 

 through spaces between parallel planes for a similar rate 

 and pressure gradient, he estimated a distance between 

 planes of 2.04^t. Since these demands of pore size are 

 obviously larger than can be expected in cell walls, he 

 concluded that part of the flow must take place through 

 the lumen of the phloem cell. Steward and Priestley 

 (1932) have given a clear presentation of the weaknesses 

 in Crafts' earlier calculations of the resistances to be met 

 in flow through phloem walls. 



By the time he wrote his second paper, therefore, Crafts 

 was beginning to appreciate the difficulties attendant on a 

 flow restricted to cell walls and suggested that part of the 

 flow might take place through the lumen. In this paper 

 he reports an apparently unobstructed flow from the 

 phloem of the cut peduncles of cucumber fruits, even when 

 the sieve plates are covered with thick callous plates. There 

 are also interesting calculations on rates of flow and 

 valuable material on the anatomy of Cucurbita phloem. 



When he finally realized the impossibly great pressures 

 necessary to force solutions along through the walls and 

 when further investigations led him to conclude that sieve- 

 tube protoplasm is completely permeable, he modified his 

 first proposal and suggested (1933) that the entire cross 

 section of the phloem would act as the channel of transport. 

 In this later paper he proposes that the principal path of 

 transport is the lumen of the sieve tubes, while the flow 

 from cell to cell may take place not through sieve pores or 

 protoplasmic strands but by filtration across any end or 

 side walls of the sieve tube that might intervene. He 

 assumes that the sieve-tube protoplasm is different from 

 that of other cells in that it is completely permeable. The 

 basis for this assumption is that in sections mounted for 

 observation the mature sieve tubes do not behave as young 

 sieve tubes or other living cells in that they fail to accumu- 

 late neutral red, stain easily with analin blue, and are also 

 incapable of being plasmolyzed. It is possible, however, 



