TRANSACTIONS 01' THE SECTIONS. 87 



ing away very far from the original stock. To the former their seems no limit ; 

 but the latter is stopped by the increasing uuproductiveuess and unhealthiness of 

 the individuals, liy the susceptibility to disease, and the tendency to revert to the 

 original type. So that increasing departure requires greatly increasing care ,• and 

 ■we do not know that any amount of care and time would be sufficient to produce 

 what might fairly be called a new species. The bringing about any marked 

 change by nature's selection is shown to be verj^ hard of proof, and ha.s opposed 

 to its probability the fact that the members of a species which are most unlike 

 have the greatest tendency to pair and are the most fertile ; so that we have here, 

 in addition to the ready reversion of modified breeds to the original stock, a law 

 by which the growth or perpetuation of ]5eenliarities is prevented and a constancy 

 given to the characters of the species. This law is more striking from its conti-ast 

 with the bar that exists to the pairing of different species and the infertility of 

 hybrids. Within a given range dissimilarity promotes fertility ; beyond that 

 range it is incompatible with it. 



These and other considerations have always inclined me to the opinion that 

 modifications of animal t}'pe, occurring in nature, are more likely to be the result 

 of external influences operating upon successive generations, influencing their 

 development, their growtli, and their maturity, than of " natural selection " and 

 "struggle for existence." But greater eifects of these and other similar agencies 

 must be shown before we ought to admit even the reasonable probability of their 

 power to work out the great changes that have been attributed to them. 



In pondering over the definiteness of animal t^^es, so marvellously elaborated 

 fi'om a simple fonn, their slight variability through long periods, the clear manner 

 in which they, many of them at least, are worked out from one another, and which 

 increasing investigation seems to render more and more apparent, the prospect of 

 proving that they are educed from one another ^3y any of the hitherto supposed 

 processes seems to gi-ow more and more distant, and the feeling arises that there 

 must be some other law at work -which has escaped our detection. 



We are familiarized with the fact that in the inorganic world combinations take 

 place only in certain definite proportions ; for instance, that oxygen unites with 

 nitrogen in one proportion to make nitrous oxide in a second proportion, a mul- 

 tiple of the first to make nitric oxide, and so on to the fifth proportion or multiple, 

 which gives iiitric acid, and that between them five several fixed proportions as 

 combinations take place. So that the resultants of these and other similar com- 

 binations (the inorganic species, as we msij call them) are remarkably constant 

 and fixed in their characters ; each has its one fomi, as in the case of crystal, of 

 chloride of sodium, or sulphate of magnesia, which may be broken down or dis- 

 solved, but which cannot be modified or made to approach, still less to pass into, 

 any other form. 



ilay there not be something analogous (some corresponding law of combining 

 proportion) presiding o\ev living matter, educing the various forms, fixing their 

 characters, gi\'ing tliem constancy, in fact evolving and fixing the species and pre- 

 venting their transmutation. 



It will be understood that I am not speaking of the combining propoi-tions of the 

 elements in the several animal tissues, which we know, or have every reason to 

 believe, is as fixed as in ordinary inorganic matter, though the combinations are 

 more complex and the forumlfe are, in consequence, harder to work out. I speak 

 now not of this, but of something comparable with this and suggested by this, 

 operating not upon individual particles, but on masses, regulating not the chemical 

 composition and form and feature of the tissues, but the form and features of the 

 animal. As oxygen unites \vith nitrogen only in the definite multiple proportions 

 represented by the figures 1, 2, .3, 4, 5, and under certain circumstances, producing 

 in each instance a special compound unlike any other and marked off fiom the 

 nearest approaching compounds by distinctive features, and ^^^thout any interme- 

 . diate gTadations, so in the animal and vegetable world the combinations requisite 

 for evolving living beings may be regulated in a similar manner, taking place only 

 in certain fixed proportions and under certain circumstances and educing certain 

 definite forms, each of which is unlike any other and is marked off from its 

 nearest approach by clearly distinctive features and without intermediate grada- 



