Crafts et al. —10— Water in Plants 



This structure supports the view, obtained from x-ray analysis of ice, 

 that each oxygen atom is surrounded by four atoms of hydrogen and that 

 each hydrogen atom lies symmetrically between two atoms of oxygen. A 

 clear picture of such structure can only be obtained from a three dimen- 

 sional view of the crystal lattice model. Such a view shows that the term 

 ice molecule has no meaning, the atom being the only real unit of structure. 

 Pennycuick pictures the hexagonal units in ice fabricated into a con- 

 tinuous lattice structure ; the primary and secondary valences are of equal 

 strength, making the whole structure symmetrical. 



In liquid water HOH chains would be irregular and distorted because 

 of tetrahedral structure; if the ends of a chain combine and neutralize each 

 other, the resulting ring would be relatively inactive {cf. Figure 4), Such 

 rings should play important parts in the structure of water and ice, and the 

 tetrahedral form of the water molecule led Pennycuick to conclude that 

 such rings in water must contain six oxygen nuclei. 



In contrast with the crystal lattice structure of the solid, with zero re- 

 sultant field around each atom, in the liquid the molecule preserves its in- 

 dividuality and the system lacks stable structure. With definite resultant 

 molecular fields aggregation would occur, the whole liquid forming a three- 

 dimensional, continually-changing network with linkages of varying 

 strength in which a single linkage could exist only momentarily. Every 

 normal liquid must possess this type of association. Non-polar or normal 

 liquids, having symmetrical arrangement of valence forces, would not form 

 definite polymers, all molecules being constantly in a state of flux. Mole- 

 cules of polar or associated liquids, having asymmetric fields, would tend to 

 combine to form stable groups, and these {i.e., Figure 4) would behave 

 like solute molecules moving as independent units through the remainder 

 of the liquid which is normal in its behavior. This view of Pennycuick 

 differs from the older one that water is a mixture of mono-, di-, and tri- 

 hydrol, each having a statistically average existence expressible as a con- 

 stant value. 



HO-H:0:HO:=^ 



• • • • • • 



H H H 



H:0:H 



• • 



H 



++ 



• • • • 



H H 



Fig. 5. — Diagram indicating the self-ionization of a water chain. (From Penny- 

 cuick, 1928). 



If water is a mixture of relatively stable hexagonal molecules in a nor- 

 mal solvent, it may represent an intermediate step between vapor and the 

 hexagonal crystal structure of ice. The auxiliary fields of the molecules 

 are presumably strong enough to build as far as the (HOH)^ ring but the 

 molecular energies are sufficiently strong to prevent coordination into a 

 lattice. The fraction of water polymerized would vary with temperature 

 and Pennycuick considered the values of Richards and his co-workers 

 as most probable, i.e., about 28 per cent as polyhydrol at 20° C. 



In terms of the above suggestions, Pennycuick proposed that when 

 ice melts there is a loosening of the structure and a breaking of the rings, 

 resulting in a contraction of volume or an increase in the density of the 

 liquid. On the other hand, the increase of molecular energy with rise in 



