374 NATURAL SCIENCE. May. 



into chemical problems. Chemical compounds are themselves 

 discontinuous, and if the chemical nature of a substance be altered 

 at all, it will be changed either inappreciably or completely. 



After this introductory matter, which alone can be treated of at 

 any length in a short account of Mr. Bateson's book, the author 

 passes directly to the exposition of the large number of cases of 

 meristic variation he has examined. These are arranged under 

 various natural headings, and are described under consecutive 

 numbers, so that future additions and references can be made very 

 simply. It is impossible to exaggerate the skill and care with which 

 this has been done, and there can be no question that a most excellent 

 introduction to the study of variation has been written. 



Although it is quite certain that a just appreciation of the value 

 of the general conclusions Mr, Bateson foreshadows can be reached only 

 after a long study, not only of Mr. Bateson's careful descriptions, but of 

 actual specimens, a few of the many considerations suggested by the book 

 maybe set down. The more one studies his book, the more one is struck 

 by the always present and occasionally avowed placing in juxtaposition 

 of the discontinuity in species and the discontinuity in variation. In 

 more than one place Mr. Bateson suggests definitely that the former 

 finds its explanation in the latter. Most will agree that the author 

 over-estimates the value of his suggestion. Take first the facts upon 

 which the so-called discontinuity in species depend. Although, in many 

 cases, it is merely an assumption of palaeontologists that those animals 

 which can be arranged in a structural series necessarily constitute a 

 series in descent, yet in many cases — as, for instance, in Ammonites 

 — series of descent upon almost undoubted evidence have been shown 

 to exist ; and these series show that the evolution of species took 

 place, not per saltum, but by continuous modification. We do not 

 expect to find complete records of continuous descent ; but the 

 imperfection of the record is necessarily so great, that unless descent 

 by simply-graded modifications were the ordinary history, it would be 

 almost inconceivable that so close simulation of such an occurrence 

 as has been shown to exist could exist. Moreover, the palaeontological 

 record explains so many cases of geographical distribution, that it 

 seems quite unnecessary to make so much difficulty about the present 

 occurrence of most nearly-allied species in the same localities. 

 Nothing has been established more certainly than that existing 

 distribution depends upon so many slight changes in the immediate, 

 and in the remote past, that the present occurrence of species is 

 absolutely no guide to their original habitats. Temporary geographical 

 isolations of parts of species repeatedly occur, and as repeatedly dis- 

 appear. We have much evidence that goes to show that the great 

 land areas of the world have remained comparatively constant in 

 position, but that there have been repeated changes from continents 

 to archipelagoes and back again to continents. We have direct 

 evidence as to the effect of geographical separation in producing 

 divergence. Putting the two together, we have good reason to sup- 

 pose that groups that have become divergent when separated, have 

 been thrown together again. Anyone who is familiar with the 

 physical changes of a country side in the very few years that cover 

 the observations of one individual, must know of changes in the 

 distribution of the most common animals and plants. This knowledge 

 may not be, and probably is not, enough to solve the problem of the 

 co-existence of different forms closely allied ; but it is enough to give 

 pause before finding in the co-existence of such species a problem 

 that can be solved only by the assumption of a discontinuous origin. 



