THE NATURE OF THE GENE 363 



between similar sets of genes. It is possible, however, that the London- 

 van der Waal's forces would act in such a way as to appear as specific 

 attractions. The magnitude of these forces depends on the difference 

 in refractive index between the particles and the medium. Now, the 

 presence of chromomeres probably indicates that a chromosome at 

 zygotene consists of a linear array of more and less refractive particles, 

 and if this is true the attraction between two sections of chromosome 

 will be greatest if the sections are homologous so that similar particles 

 can fit side by side. A homologous pairing arrived at by chance in one 

 region would therefore be expected to extend along the chromosome by 

 a mechanism like that of a zip fastener. Some indication that these 

 forces are really involved in pairing might be found if regions of chromo- 

 some which become precociously condensed in zygotene, and therefore 

 have a higher refractive index than the rest of the chromosome, were 

 to pair precociously. But the invocation of the London-van der Waal's 

 forces to explain pairing can only be regarded as the simplest hypothesis, 

 and is still in need of rigorous checking by specially designed observa- 

 tions and experiments. 



For the forces of the external mechanics, operating over greater 

 distances, it is probably necessary to look outside the chromosomes 

 themselves to the spindle. The physical nature of the spindle is now 

 becoming fairly clear. On fixation it coagulates as a "spindle-shaped" 

 body Hke two cones united by their bases; internally it has a structure 

 of long fibres. Micro-dissection of living spindles^ shows that these 

 fibres do not exist as such in life; the micro-needle can be moved 

 through them without displacing the chromosomes to which they are 

 apparently attached. On the other hand, the formation of fibres on 

 fixation is undoubtedly an expression of some real factor in the consti- 

 tution of the spindle. This factor can hardly be a magnetic, electrical or 

 diffusion field, since spindles can be caused to bend into quite sharp 

 U-shaped curves. The field can only be a material one, and one may 

 probably assume that the spindle is a region in which elongated particles 

 lie parallel to one another in an arrangement like that in a nematic 

 Hquid crystal. 2 Positive evidence for this suggestion is provided by the 

 behaviour of the spindle in polarized light.^ Moreover, the spindle is the 

 equiUbrium shape for a mass of orientated elongated particles floating 

 in a Hquid medium with wliich it is immiscible, while the radiating 



^ Chambers 1925, Wada 1935. 



^ For the physics of liquid crystals, cf. Bragg 1933. 



^ Schmidt 1936, 1937. 



