606 



SCIENCE. 



[N. S. Vol. XI. No. 277. 



CH,-CH— CH— CH3, CH— CH<^^|. 



Sucli formulas are not intended to repre- 

 sent the actual form of the molecule ; they 

 are schematic merely ; they are not fanci- 

 ful, but are based on innumerable experi- 

 ments, which do not concern us here. The 

 innumerable facts of organic chemistry tally 

 so well with the assumption of chemical 

 units or atoms, linked together in definite 

 ways, as to give to the hypothesis a very 

 great degree of probability. 



Some atoms can manifest a different 

 valency according to circumstances ; thus 

 iron forms two chlorides, FeClj and FeClj, 

 in the former of which the iron atom 

 is believed, for good reasons, to have but 

 two valences (Fe"), and in the latter three 

 (Fe'"). 



Of the cause of valency we know nothing, 

 and in addition to what I have said we 

 know but little, except that there is an 

 intimate relation between the valences and 

 the power of carrying electric charges, and 

 that (in some cases at least) there is a defi- 

 nite geometrical relation between the va- 

 lences. 



If a salt, for example, sodium chloride, 

 be dissolved in water, and an electric cur- 

 rent be passed through the solution, the 

 sodium atoms move to the negative pole, 

 carrying charges of positive electricity, and 

 the chlorine atoms to the positive pole, 

 carrying negative electricity. These elec- 

 trically charged atoms (or groups of atoms) 

 are termed ions. If the chlorides of iron 

 be similarly treated, a similar result follows, 

 the iron transporting positive electricity. 

 There is this difference, however, that the 

 divalent iron atom in FeCl^ transports but 

 two units of electricity, while the trivalent 

 atom in FeC]3 carries three. More complex 

 molecules are frequently broken up into a 

 mixture of simple and complex ions, the 

 latter carrying a unit of electricity for each 

 free valence, thus K^SO^ gives two positive 



K ions and the negative ion = SO^, which 

 carries two units ; with regard to the 

 amount of electricity carried the valences 

 are therefore equivalent. What the cause 

 of the relation between valency and electri- 

 cal phenomena is, we do not know. 



There is still another property of valency 

 which has been discovered in recent years, 

 and which is of the highest importance. It 

 may be asked whether in the case of the 

 carbon atom, for example, the division of 

 the attractive power into four valences im- 

 plies action in four distinct and fixed direc- 

 tions, that is, whether the atom possesses a 

 sort of polarity, or whether the action may 

 be in any direction. Thanks to the recent 

 labors of organic chemists, it now seems 

 tolerably certain that the carbon atom tends 

 to exert its attraction in four distinct and 

 tolerably fixed directions, rather than in 

 all directions equally. 



It has long been known that there exist 

 certain pairs of organic compounds which 

 have identical composition and molecular 

 weight, which show identical chemical be- 

 havior and which agree in all physical 

 properties except two. These bodies can- 

 not be regarded as isomeric in the ordinary 

 sense, as their structural formulas as usually 

 written, are identical. The two respects in 

 which they differ are these : in solution one 

 rotates the plane of polarization of light to 

 the right, the other to the left ; when crys- 

 tallized, they frequently show hemihedral 

 faces, differing only in this, that the crys- 

 tals of the one cannot be brought to coin- 

 cide with those of the other, but are as an 

 unsj'mmetrical body and its reflected image 

 in a mirror. I have said that the ordinary 

 structural formulas are schematic merely, 

 they do not claim to show the actual rela- 

 tion of the atoms in space. Two mole- 

 cules, whose geometrical forms are identical 

 except in being right- and left-handed or as 

 object and reflection, would be represented 

 by the same structural formula, and would 



