APPENDIX A 305 



case the standard deviation falls, while the correlation coefficient 

 rises, when they are both removed. It will be observed that, 

 while gravitation (y) has less effect than compression (/3) upon the 

 angles B and C, the reverse is the case with the angle A. We 

 may be able to find a reason for this later on. 



There is one point worth noticing. It is quite clear that 

 gravity is not indispensable for the development of a grey 

 crescent and plane of symmetry, though it is true that the position 

 of this plane may be affected by gravity even in the short interval 

 that elapses before the egg turns over. 



The values for the compressed eggs with horizontal axes (a) 

 compare fairly well with those previously obtained, except in the 

 case of the plane of symmetry and the first furrow. In the 

 former series the latter tended either to coincide with or to lie 

 at right angles to the former. In the present series this is not 

 the case. This difference is probably to be attributed to the 

 fact that many of the eggs in the first series must have been 

 placed on the slide with the white pole upwards : possibly also 

 the ' compression ' was greater then than now. 



It is fortunate that the same data enable us to study exactly 

 the relation between the first furrow and the plane of symmetry 

 on the one hand, and the direction of ' compression ' and of the 

 gravitation symmetry plane on the other. It must be remem- 

 bered that these two are at right angles to one another. 



Consider first the first furrow. 



(a) When the eggs are close but the axis horizontal the first 

 furrow tends to lie at right angles to the slide, that is, in the 

 direction of compression, but at right angles to the gravitation 

 symmetry plane. (<r = 38-16 + -69.) 



(/3) When the eggs are close but the axis vertical this tendency 

 is not quite so marked. (0- = 46-67 + -71.) 



(y) When the eggs are spaced and the axis horizontal it is 

 still there, but slight, (a- = 49-32 1 -40.) 



(8) When the eggs are spaced and the axis vertical the 

 direction of the first furrow is random. (a = 52-76 + 1-17.) 



We may conclude, therefore, that the first furrow tends to lie 

 in the direction of the ' compression' and at right angles to the 

 plane of gravitation symmetry. The latter tendency, we know, 

 exists in forcibly inverted eggs, together with a tendency to lie 

 in the plane of symmetry and at 45 to it (above, p. 84). 

 Pressure experiments also show that division is in the direction 

 of pressure (p. 34 sqq.). 



The direction taken up by the plane of symmetry under these 

 different circumstances is quite distinct from that of the first 

 furrow. It appears to be determined in the first instance by 



