PROGRESS MEANS EARLY DEVELOPMENT 527 



another. Hence if OP be the type-path, we may take as representative 

 individual paths OKR and OMQ, lying near OP, but not as a rule 

 crossing it -(X.). 



34. Earlier development. Now this has an important consequence. 

 Instead of drawing, as hitherto, ordinates such as QPRT, parallel to the 

 OY axis, draw one NMLK parallel to the OT axis. Then NM or OV 

 represent the time at which the child, OQ, has the organ under consideration 

 of the same size (MV or LX) as the parent at the time OX. The child, 

 OQ, reaches this stage earlier than the parent ; and since the same is true 

 throughout the curve we see that this child passes through all the successive 

 stages (in regard to the organ in question) earlier than its parent. 

 Similarly the child, OR, develops more slowly than its parent, throughout. 

 We may say generally that individuals strong in any organ develop early ; 

 those weak develop late (XI.). 



35. Relation of organs. But we have so far considered one organ 

 alone, whereas the characteristics of any animal depend on the relation of 

 its parts small head and long neck 



in a giraffe : large head and short 

 neck in a lion. What is true of an 

 isolated organ may not be true of a 

 combination, and we must consider 

 how the relationship of one to another 

 enters into the questions before us. 

 In the first instance, take any two 

 organs such as length of leg and 

 weight of brain : and, dropping any 

 question of time for the moment, let 

 us make a diagram to represent their 

 relationship, setting off the length of 

 leg along OZ, and the weight of brain along OV. Thus the point P repre- 

 sents the fact that when the length of leg is 4 feet, the weight of brain is 

 3 ounces or pounds, or any other suitable unit. Previously (at Q) the 



length of leg was 2 feet, and the weight 

 of brain was i oz. or Ib. Since the 

 dimensions are all zero at first, the path 

 PQ will start from O, just as that repre- 

 senting the relation bet ween- any organ 

 (say Y) and the time. 



36. Now it will easily be seen that 

 most of the considerations already 

 advanced in the case of the paths for 

 Z the growth Y of an organ, with the 

 time T, will apply equally well to the 

 path for the growth of one organ 

 relatively to another. If OQP is the path for a parent, and Oqp that of 

 a child, the facts that children roughly resemble parents, but that they 

 may differ from them in either direction, and so on, still apply to these 

 new paths. Hence we get the same rules about sexual and ancestral 



FIG. 15. 



