i894. SOME NEW BOOKS. 373 



have shown, causes, when complete, two individuals, when incom- 

 plete, a double monster. 



To make plain what he means by discontinuity in variation, Mr. 

 Bateson gives a number of instances of discontinuity in substantive 

 variation. In the case of 583 mature male earwigs collected at the 

 same time, in the same locality, it was found that the length of the 

 forceps varied from two and a half to nine millimetres. In the most 

 common cases of variation of this kind, it is found that, when a curve 

 is drawn of which the ordinates represent the numbers of individuals, 

 and the abscissae the lengths, that the curve is a simple one, corres- 

 ponding to what is called a curve of error. In other words, individuals 

 with a mean length are the most common, individuals with extreme 

 lengths, above and below the average, are most rare. In Mr. Bateson's 

 earwigs there were two curves of error, two common kinds of earwigs. 

 About 120 had a length of three and a half n:m., about 90 of seven 

 mm., and the numbers tapered off to the extremes and to the mean. 



Sudden variations of colour are very common, and as these are 

 generally from one colour directly to another colour without inter- 

 mediate shades, it is clear that colour-variation may be discontinuous. 

 Other common discontinuous variations are hairless mice or rats, bull- 

 dog headed fish, and so forth. As typical examples of what he means 

 by discontinuity in meristic variation, Mr. Bateson takes two cases. The 

 first is where a flower — for instance a tulip — has all the parts in 

 multiples of four, instead of in multiples of three. The change is 

 complete, the division into four and the resulting symmetry is as 

 perfect as the normal division into three. The second case is that of 

 the tarsus of the leg of a blackbeetle. Usually this consists of five 

 joints, but as a common variation the tarsus of one side may consist 

 only of four. But in this case, as in the tulip, the new symmetry is 

 complete. The four-jointed tarsus is of the usual length, and it is as 

 perfectly shaped, and apparently as workmanlike a production, as the 

 normal form. These two cases show what Mr. Bateson's extended 

 observations clearly establish. Variations so massive as to be 

 clearly discontinuous with the normal form, show for the most part 

 no trace of monstrosity or irregularity, but are so clearly co-ordinated 

 to their function, and to the general symmetry of the body, that only 

 by actual knowledge can the variation be detected. If a number 

 of examples of the variation were placed in the hands of an observer, 

 who, though otherwise competent, were ignorant of the particular 

 animals or plants, he would detect nothing unusual in the variations 

 given him. Speaking generally, Mr. Bateson finds that what Galton 

 called positions of organic stability occur in variations. Just as in 

 species there is a mean form, on either side of which slight irregularities 

 taper off, so in most variations there is a normal or mean form, round 

 which slight irregularities of the variation cluster. This suggests to 

 him that the discontinuity in species may be an expression of the 

 discontinuity of variations. Of course he puts forward this view, not so 

 much as an actual solution, as a direction in which the study of 

 variation may profitably proceed. He suggests, in fact, that it is worth 

 while enquiring whether the definiteness of animal and plant types be 

 not due to the physical limitations of variation rather than to 

 adaptation to environment. In connection with this, he draws 

 attention to the obvious fact that as growth depends upon cell- 

 division, many of the problems of organic structure may resolve them- 

 selves into problems of the physiology of division. Similarly, in 

 questions of qualitative change, questions of what he calls substantive 

 variation, many of the problems will resolve themselves ultimately 



