naturally be expected to spread, in time, through the whole mass" (p. 40).
But how is the transmission of acquired characters thereby rendered less mysterious? Supposing an elephant exercises his trunk to such an extent that the muscles enlarge beyond the ordinary, do the molecular changes in this case differ essentially from the molecular changes which occur in the trunk of an elephant that uses his trunk to such an extent only that the muscles of it are maintained at the ordinary standard? And if they do differ, how can these molecular changes, spreading through the mass by way of many variously differentiated structures, which, being of different molecular constitutions, must be differently affected by the molecular changes, so affect the molecular constitution of the germ cell when they reach it, that it ultimately proliferates into an organism with enlarged muscles in the trunk? It must be remembered, that neither muscle cells nor any of the other kinds of tissue present in the trunk exist in the germ cell. These only appear in its very remote cell-descendants. If anything the mystery is deepened by Mr. Sedgwick's theory, and those who hold that acquired variations are transmitted will I think prefer Mr. Darwin's theory of gemmules, or Mr. Spencer's own theory of physiological units, as furnishing more probable explanations. Moreover, though Mr. Sedgwick may have established that other cells besides nerve cells are connected by fine protoplasmic processes, he has not as yet established that their nuclei are in any way affected by the connection; for example, he has not proved that the nuclei of the germ cells are in any way affected by the supposed partial continuity of their cell-bodies with the cell-bodies of adjacent somatic cells, and until this is done his researches, however interesting
