424 ELEMENTARY PROBLEMS IN PHYSIOLOGY. 



iiess. Having arrived at this point he begins to correlate the known 

 structure of the retina with wliat is required of it, and finds that the 

 number of objects which he can discriminate in the field of vision is as 

 numerous as, but not more numerous than, the parts of the retina, i.e., 

 the cones which are concerned in discriminating them. So far he has 

 no difiQculty ; but the method of correlation fails him from the moment 

 that he considers that each cbject point in the field of vision is colored, 

 and that he is able to discriminate not merely the number and relations 

 of all the object points to each other, but the color of each separately. 

 He then sees at once that each cone must possess a plurality of en- 

 dowments for which its structure affords no explanation. In other 

 words, in the minute structure of the human retina we have a mechan- 

 ism which would completely explain the picture of which I am con- 

 scious, were the objects composing it colorless, i. c, possessed of one 

 objective quality only, but it leaves us without explanation of the dif- 

 ferentiation of color. 



Similarly, if we are called upon to explain the function of a secreting 

 gland, such, e. g., as the liver, there is no difficulty in understanding 

 that inasmuch as the whole gland consists of lobules which resemble 

 each other exactly, and each lobule is likewise made up of cells which 

 are all alike, each individual cell must be capable of performing all the 

 functions of the whole organ. But when by exact experiment we learn 

 that the liver possesses not one function but many,-e-when we know 

 that it is a storehouse for animal starch, and that each cell possesses 

 the power of separating waste coloring matter from the blood, and of 

 manufacturing several kinds of crystallizable products, some of which 

 it sends in one direction and others in the opposite, we find again that 

 the correlation method fails us, and that all that our knowledge of the 

 minute structure has done for us is to set before us a question which 

 though elementary, we are quite unable to answer. 



By multiplying examples of the same kind, we should in each case 

 come to the same issue, namely, jjlurality of function ivith imiti/ of 

 structure, the unity being represented by a simple structural element — 

 be it retinal cone or cell — possessed of numerous endowments. When- 

 ever this |)oint is arrived at in any investigation, structure must for the 

 moment cease to be our guide, and in general two courses or alterna- 

 tives are open to us. One is to fall back on that worn-out Dexis ex ma- 

 china — protoplasm, as if it afforded a sufficient explanation of everything 

 which cannot be explained otherwise, and accordingly to defer the con- 

 sideration of the functions which have no demonstrable connection with 

 structure as for the present beyond the scope of investigation ; the other 

 is, retaining our hold of the fundamental principle of correlation, to 

 take the problem in reverse, i. c, to use analysis of function as a guide 

 to the ultra-microscopical analysis of structure. 



I need scarcely say that of these two courses the first is wrong, the 

 second right, for in following it we still hold to the fundamental prin- 



