376 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1933 



At A the mesoderm {Msd) is seen growing inward as a hollow 

 fold on each side of the blastopore where the ectoderm was first 

 continuous with the endoderm. Typically, the mesoderm forms a 

 series of paired pouches, called coelomic sacs (fioeV)^ corresponding 

 with the segments of an annelid or an arthropod, for the body of 

 these animals behind the mouth is always divided into consecutive 

 parts, or segments (but segmentation is a topic we must omit from 

 the present history). The mesodermal layers grow upward around 

 the alimentary canal as the blastopore closes (B) ; their cavities 

 {(Joel) may remain distinct, but in the arthropods the coelomic 

 walls partly break down and the contained spaces unite with the 

 surrounding blood spaces {He) to form the definitive hody cavity 

 (C, BC). The outer and inner layers of the mesoderm, however, 

 form respectively a lining to the body wall {soniatopleure^ EMsd), 

 and a covering around the alimentary canal {splanchnopleure, I Msd). 



The foregoing sketch gives probably an approximately true ac- 

 count of the historic method of mesoderm formation as it can be 

 deciphered here and there from the incomplete and modified stories 

 related in embryonic development. As we have seen, the embrj'o 

 is always prone to get results by short cuts and adaptations to suit 

 its own convenience, and therefore, while the embryo is commonly 

 said in its development to repeat the history of its race, this state- 

 ment is far from being literally true in most cases. The embryo, 

 however, is not deficient in " biological memory " ; it adheres closely 

 to the general scheme and sequence of historical events, but we must 

 consider the fact that it is usually shut up inside an egg, and is 

 thus cut off from the outside world by an egg shell. Naturally, 

 therefore, it cannot carry on its life functions in the way its free- 

 living ancestors did, and as a consequence it has to adapt many 

 events in its development to the way of living that is forced upon it. 

 Then too, the embryo must complete its development in a very short 

 time, often in a few weeks or a few days in the case of insects, while 

 its ancestors took millions of years to cover the same ground. 

 Hence, the embryo cannot be bothered with details, and it must 

 attain its ends by expedient processes. Furthermore, the farther 

 its immediate progenitors have departed from the structure of their 

 early ancestors, the more difficult becomes the task of recapitulation, 

 and the less likely is the embryo to follow the traditional path. 



Let us now consider specifically the case of the insect embryo, 

 some of its developmental problems, and particularly the way in 

 which it builds up a conventional alimentary canal in spite of the 

 roundabout methods it is forced to adopt. The egg of an insect is 

 like that of any other animal in that it is a single cell, but the eggs 

 of most insects are relatively very large because each contains an 



