There are essentially two conceptions of early development of somatic movements in mammalian embryos. One group of investigators, Coghill, Angulo y Gonzáles, Hooker and their colleagues,1 have expressed the belief that they have demonstrated the genesis of reflex activities by a process of "individuation" from an already integrated total reaction, or pattern; in other words, that more or less discrete movements are not the primary units but that these break off, as it were, from some sort of a mass movement that constitutes the basic pattern of behavior. Another group of investigators, Carmichael, Windle and their associates,2 have expressed the opposing view that relatively simple activities manifest themselves first and that these become integrated during further growth in the nervous system, so that the more complex reactions of older fetuses are built up from the less complicated ones of the younger. The issue is clearly drawn, and it