During the next few years Roux continued his studies on functional adaptation,[472] and at the same time made a new departure by inaugurating, almost contemporaneously with the physiologist Pflueger, the study of experimental embryology. Isolated observations had previously been made upon the development of single blastomeres or parts of blastulae, by Haeckel and Chun for instance,[473] but Roux[474] and Pflueger[475] were the first to investigate the subject systematically, choosing for their work the egg of the frog.[476] Roux continued for many years to follow up this line of work.[477] In 1890 he drew up a programme and manifesto[478] of _Entwicklungsmechanik_ as "an anatomical science of the future," and in 1895 he founded the famous _Archiv fuer Entwicklungsmechanik_,[479] publishing in the same year the two large volumes of his collected papers,[480] of which the first volume dealt with functional adaptation, the second with experimental embryology. His subsequent work includes several important general papers;[481] besides a number of special memoirs dealing with the factors of development, and with his original subject, functional adaptation.[482] In our sketch of his views we shall have occasion to refer particularly to his publications of 1881, 1895 (the _Einleitung_), 1902, 1905, and 1910. Although Roux's biological philosophy is out-and-out mechanistic, he yet recognises the difficulty, even the impossibility, of straightway reducing development to the physico-chemical level. He tries to steer a course midway between the simplicist conceptions of the materialists and the "metaphysics" of the neo-vitalist school, which the experimental study of development and regeneration soon brought into being. In 1895 he writes:--"The too simple mechanistic conception on the one hand, and the metaphysical conception on the other represent the Scylla and Charybdis, between which to sail is indeed difficult, and so far by few satisfactorily accomplished; it cannot be denied that with the increase of knowledge the seduction of the second has lately notably increased" (p. 23).