When the contents of the egg are displaced by pressure the result will be determined by the loca­tion of the main mass of the intestine-forming material; where the main mass of this body is located the invagina­tion of the intestine will take place. In his earlier work Driesch assumed from pressure experi­ments that the egg had a great power of “regula­tion.” In a later paper[122] he expressed to a large extent his agreement with Boveri who denied this power of “regula­tion” and showed that the existence of the structure of the egg—i. e., a division into three strata, one forming the ectoderm, the second the entoderm, and the third the mesoderm—was sufficient to explain the various phenomena of apparent “regula­tion.” Driesch’s idea of a regula­tion in this case has often been used to insist upon the non-explicability of the phenomena of development from a purely physico­chemical viewpoint. It is, therefore, only fair to point out that Boveri[123] has furnished the facts for a simpler explana­tion, which seems to have escaped the notice of antimechanists.[124]

The objection may be raised that in accepting Boveri’s facts and interpreta­tion we pushed the miracle only one step farther and that we now have to explain the origin of the structure in the unfertilized egg. This Boveri has done by showing that the egg grows from the wall of the ovary and that that part of the egg which is connected with the wall of the ovary gives rise to the ectoderm layer, while the opposite part gives rise to the mesenchyme and the intestine. This shows a connec­tion between the orienta­tion of the egg in the wall of the ovary and its stratifica­tion. While this does not solve the problem of stratifica­tion in the egg it gives the clue to its solu­tion.

The ultimate origin of stratifica­tion probably goes back to the fact of the presence of watery and water-immiscible substances, such as fats. The experi­ments by Beutner and the writer have shown that the electromotive forces which are observed in living tissues originate at the boundaries between a watery and a water-immiscible phase, like oleic acid or lecithin.[125] In his earlier writings[126] the writer had thought that the colloids had special significance and this idea seems to prevail today; but the actual observa­tions have shown that the phase boundary fat-water is of greater importance. Needless to say the fats if not present in the cell from the beginning can be formed in the metabolism.

4. All the “regulation” in the egg is of a purely physico­chemical character; it consists essentially of a flow of material. If this idea is correct, the apparent power of “regula­tion” of the blasto­meres should differ according to the degree of fluidity and the possibility of different layers separating, and this assump­tion is apparently supported by facts. The first plane of segmenta­tion of the egg is usually the plane of symmetry of the later organism and where the degree of fluidity is less than in the sea-urchin egg, a separa­tion of the two first blasto­meres should easily result in the forma­tion of two half-embryos instead of two whole embryos.

This is the case for the frog’s egg as Roux showed in a classical experi­ment. Roux destroyed one of the two first cleavage cells of a frog’s egg with a hot needle and found that as a rule the surviving cell developed into only a half-embryo.[127] The frog’s egg consists of two substances, a lighter one which is on top and a heavier one below. Although viscous, the two substances are not too viscous to prevent a flow if the egg is turned upside down. O. Schultze found that if a normal egg is turned upside down in the two-cell stage and held in that posi­tion, two full embryos arise, one from each of the two blasto­meres. Through the flow of the lighter liquid in the egg upwards the two halves of the protoplasm on top become separated and develop independently into two whole embryos instead of into two half-embryos. In Roux’s experi­ment this flow of protoplasm was avoided. Morgan showed that if Roux’s experi­ment is repeated with the modifica­tion that the egg is put upside down after the destruc­tion of the one cell, the intact cell will give rise not to a half but to a whole embryo.[128] These experi­ments prove that each of the first two cleavage cells of the frog’s egg represents one-half of the embryo and that a whole embryo can develop from each half only when a redistribu­tion of material takes place, which in the egg of the frog can be brought about by gravita­tion since the egg consists of a lighter and a heavier mass.

When, therefore, in the egg of the sea urchin each of the first two blasto­meres naturally gives rise to a whole embryo it is due to a greater degree of fluidity of the protoplasm and not to a lack of preforma­tion of the embryo in the cytoplasm. This idea is confirmed by the observa­tions on the egg of Ctenophores whose cytoplasm seems to be more solid than that of most other eggs. Chun found that the isolated blastomere of the first cell division produced a half-larva, possessing only four instead of the eight locomotor plates of the normal animal.

It seems that in the egg of molluscs, also, the simple symmetry rela­tions of the body are already preformed. It is well known that there are shells of snails which turn to the right while others turn in the opposite direc­tion. The shells of Lymnæus turn to the right, those of Planorbis to the left. It was observed by Crampton[129], Kofoid, and Conklin that the eggs of right-wound snails do not segment in a symmetrical, but in a spiral, order, and that in left-handed snails the direc­tion of the spiral segmenta­tion is the reverse of that of the segmenta­tion in the right-handed snails. Conklin was able to show that the asymmetrical spiral structure is already preformed in the egg before cleavage. The asymmetry of the body in snails is therefore already preformed in the egg.[130]

E. B. Wilson[131] has found a marked differentia­tion in the eggs of some annelids and molluscs. He isolated the first two blasto­meres of the egg of Lanice, an Annelid. These two blasto­meres are somewhat different in size; from the larger one of the first two blasto­meres, the segmented trunk of the worm originates. Wilson found that

when either cell of the two-cell stage is destroyed, the remaining cell segments as if it still formed a part of an entire embryo.[132] The later development of the two cells differs in an essential respect, and in accordance with what we should expect from a study of the normal development. The posterior cell develops into a segmented larva with a prototroch, an asymmetrical pre-trochal or head region, and a nearly typical metameric seta-bearing trunk region, the active movements of which show that the muscles are normally developed. The pre-trochal or head region bears an apical organ, but is more or less asymmetrical, and, in every case observed, but a single eye was present, whereas the normal larva has two symmetrically placed eyes. The development of the anterior cell contrasts sharply with that of the posterior. This embryo likewise produces a prototroch and a pre-trochal region, with an apical organ, but produces no post-trochal region, develops no trunk or setæ, and does not become metameric. Except for the presence of an apical organ, these anterior embryos are similar in their general features to the corresponding ones obtained in Dentalium. None of the individuals observed developed a definite eye, though one of them bore a somewhat vague pigment spot.

This result shows that from the beginning of development the material for the trunk region is mainly localized in the posterior cell; and, furthermore, that this material is essential for the development of the metameric structure. The development of this animal is, therefore, to this extent, at least, a mosaic work from the first cleavage onward—a result that is exactly parallel to that which I earlier reached in Dentalium, where I was able to show that the posterior cell contains the material for the mesoblast, the foot, and the shell; while the anterior cell lacks this material. I did not succeed in determining whether, as in Dentalium, this early localiza­tion in Lanice pre-exists in the unsegmented egg. The fact that the larva from the posterior cell develops but a single eye, suggests the possibility that each of the first two cells may be already specified for the forma­tion of one eye; but this interpreta­tion remains doubtful from the fact that the larva from the anterior cell did not, in the five or six cases observed, produce any eye.