242 MR. MO RLE Y ROBERTS ON TUE FCNCTION OF 



satisfaction for the instinct of workmanship. Human ornament 

 is in fact strongly iiomologous, if we may use that word here, 

 witli the beauty of very energetic birds, wlio cairy out by virtue 

 of their free energy the extension of sti-uctures and colours 

 already existing in their less brilliant forms. That, however, is 

 by the way. The main fact we are concerned with is that tlie 

 structure as a whole evolved through trial and error, through 

 failure and repair, through a threatened structure to a moi'B 

 complete and adequate one for increased function. In a woi-d, 

 the great origin of structure was failure after failure duly com- 

 pensated for. Is there any reason for believing that variation 

 in the structure of living ^organisms follows exactly the same 

 principle 'i Are we entitled to say that the mammal, for in- 

 stance, with all its complexity, is the result of infinite ages of 

 functional failure or disease which was met by processes of i-epair 

 and reaction ? In a word, can we speak of the evolutionary 

 value of disease, of impaired function, of disadvantageous varia- 

 tions 'i It seems possible to do so, if what is ti-ue of one structure 

 is roughly true of another. 



It may seem absurd to speak of the value of disadvantageous 

 variation, but it is no more absurd than to imply that all 

 vai'iation is advantageous because it is perjietuated. What is 

 useful at one period may be harmful at another, and embryolo- 

 gists thoroughly understand that developments useful in fcetal 

 oi' larval life may open up many dangei'S for the adult, Tlie 

 real point to be considered is whether organisms as species do 

 not vary and run great, even largely destructive, risks by an 

 increased pressure of function which, in the few that finally 

 react or whose descendants react to such stress, results at last 

 in structure that is advantageous as cdtered. The given variation 

 in itself may be a failure of what was normal function in the 

 species, and we should therefore as pathologists or physiologists 

 speak of it as disease, but if the few that recover become a new 

 species, a mended race, it is no longer disease. After many 

 generations it may be truly advantageous to individuals. Have 

 such processes occurred in the evolution of organisms as they 

 undoubtedl}^ have in the arts and social progress, where we often 

 observe political failure or organization result in ad hoc I'eaction 

 which leads to a changed social form? I have no doubt that 

 they do, and many organs in mammals, to speak only of them, 

 show it. It is, in fact, a universal principle. As beavers patch 

 up a dam when it yields or threatens to give way, so tissues, 

 organs, and societies react to threatened disaster. In no tissue 

 is this clearer, than in bone. It is true that Wolff's law only 

 deals directly with mechanical stresses, since it runs — " every 

 changa in the form and position of the bones or their function is 

 accompanied by certain definite changes in their internal archi- 

 tecture and by equally definite secondary alterations of their 

 external conformation in accordance with mathematical law;" 

 but I hope to show reasons for concluding that such a law may 



