398 AN INTRODUCTION TO MODERN GENETICS 



of individual indivisible particles, each of which is a gene, is too simple. 

 In attempting to work out a more adequate picture, one can start from 

 the fundamental fact that the chromosome is an elongated structure 

 which, whenever we can analyse it, has differences arranged in a linear 

 order along it; these differences can be detected by linkage studies, 

 chromosome structures, etc. The units, between which differences are 

 noted, may be of different sizes according to the different methods of 

 investigation; there are, in roughly descending order, inert or pre- 

 cociously condensing regions, large chromomeres, ultimate chromo- 

 meres or salivary gland chromomeres, and the units of cross-over and 

 X-ray breakage. One might symbolically represent the chromosome 

 thus: abcd'e'f'g'hijklMNOPQRSTU'V'W' where there are differences 

 on three scales, between the capitals and low^er-case letters; normal, 

 underlined and dashed letters; and finally the letters themselves. 

 The smallest units of this scheme, symbolized by the individual letters, 

 are the units of crossing-over and X-ray breakage, and probably 

 measure, as we have seen, about lOO m^ in length. 



If we view the chromosome as it were through the other end of the 

 telescope, attempting to build it up from chemical units, we arrive at 

 a somewhat similar scheme of a linear order of units of different orders 

 of magnitude. The ultimate units now are the links in a polypeptide 

 chain, with a length of only 0-334 m/x. Exactly what the larger units 

 are is more doubtful, but we have a range of possibiHties; there are 

 the periodicities along the chains,^ the repeat units out of which protein 

 crystals are built, the protein molecules such as they exist in solution, 

 and finally virus particles, all of which may be considered as providing 

 suggestions as to the kinds of imits which may be involved. These 

 units range in size nearly up to the 100 m/x which we took as an estimate 

 of the smallest units to be considered when we approached the chromo- 

 some structure from the other end. It is, then, possible to conceive of 

 the chromosome as a linear array of units, the units themselves forming 

 a hierarchy aU the way from heterochromatic and euchromatic regions, 

 some tens of thousands of m/x long, to polypeptide links only a few 

 tenths of a m/x long. 



This apparent homogeneity in the type of formal order exemplified 

 by the chromosome on different scales should not tempt one to suppose 

 that other properties may be just as easily conceived of in any of these 

 scales. For instance, it is sometimes suggested that because the nature 

 of one Unk in a polypeptide chain may chemically affect the properties 

 ^ Cf. Bergmann and Niemann 1938. 



