CELLS AND TISSUES 



17 



and Strasburger found no plasmodesms between the cells of Viscnm and 

 Cuscuta and their hosts, but in the case of graft hybrids both Buder (1911) 

 and Hume (1913) report their presence in the walls separating cells which 

 are supposed to be genetically unrelated. This seems to show that con- 

 nections may arise secondarily, although uncertainty regarding the exact 

 behavior of the protoplasts in the wounded region leaves an element of 

 doubt. They arise secondarily in the abutting walls of tyloses, according 

 to Molisch (1888). 



At present the probabilities are in favor of the view that intercellular 

 strands are both primary and secondary in origin, some of them repre- 

 senting regions in w^hich the continuity of the protoplasm has never been 

 broken and others being subsequently developed through the intervening 



Fig. 11. — Endosperm of 

 Diospyros, showing plasmo- 

 desms traversing the thick- 

 ened cell walls. {After Qui- 

 sumbing, 1925.) 



" W , 



fl) 



% 



■in^ 



■mk 



:-C? 



:>: V ■ 





.-::;.) 



6r- „ __ „ _ - 



Fig. 12. — Protoplasmic con- 

 tinuity in human mesoderm 

 tissue. (After Maurer.) 



0-j , 



afsc^ 



cell membranes. The very fact that a given area of cell wall becomes 

 enormously extended during the growth of the tissues indicates that 

 many of the pores seen at maturity must have been formed anew. It 

 is also to be remembered that at the time when pore formation would 

 necessarily occur the cell membranes are very thin and semifluid and 

 would offer little resistance to dissolution or penetration by the proto- 

 plasts. Hence the fact that cells may glide or roll over one another 

 during the early stages of development does not prove the impossibility 

 of protoplasmic continuity between them. The exceeding fineness of 

 many known plasmodesms, moreover, indicates that failure to find such 

 connections in certain tissues does not necessarily prove their absence. 

 On the other hand, proof that plasmodesms are continuous through the 

 entire wall is not so complete as desired. Jungers (1930) has recently 

 called into question the evidence for the view that plasmodesms in 



