344 EMBRYOLOGY OF INSECTS AND MYRIAPODS 



time the gastrula furrow has closed anteriorly (Fig. 3005) but is not yet 

 obliterated posteriorly by the union of the edges of the ectoderm. The 

 caudal pouch (telson) at this stage has become smaller though more 

 elongated and tubular, and into it the lower layer extends, apparently 

 without a break. 



Until now the inner layer has consisted of a longitudinal band of 

 cells, distinctly enlarged at the anterior extremity and to a lesser degree 

 at the posterior end. We may conceive of this layer as composed of 

 two lateral strips which are to form the mesoderm, a middle strand, and 

 an anterior and a posterior cell mass. These two cell masses in Diacrisia 

 are regarded as the anterior and posterior extremities of the middle 

 strand, which later break down completely, the liberated cells being 

 absorbed in the yolk. 



Segmentation of the body begins soon after the fortieth hour, and 

 soon thereafter the gnathal segments of the head become well marked; 

 the protocephalon includes the labrum, which is not yet bilobed, the 

 mouth, and the antennae. At this time the antennae are still postoral 

 in position. The stomodaeum, at the fifty-fifth hour, is feebly mdicated, 

 and opposite it is found the anterior cell mass, though there is as yet no 

 trace of the proctodaeum. Paired swellings appear on the head, the 

 thorax, and to a lesser extent on the abdomen, marking the beginning of 

 antennae, mouth parts, and thoracic and abdominal legs. As the 

 appendages lengthen, a lumen is formed in them. With formation of 

 the neural groove the cells along the median line (the middle strand) are 

 loosened and set free in the body cavity. The mesodermal strips are 

 now separated by the neural groove in the posterior part of the head, in 

 the thorax, and in the abdomen except at the apex. There is as yet no 

 trace of coelomic sacs, though the lateral margins of the mesoderm are in 

 places dorsally reflexed, the margin thus appearing two-layered for a 

 depth of two or three cells. Toward the caudal end in the last two or 

 three segments is found the posterior cell mass several layers deep and 

 with no longitudinal median interruption. Neither is there any inter- 

 ruption at this stage that can be interpreted as a division between the 

 lateral mesoderm and the posterior cell mass. 



In the part of the head in front of the stomodaeal invagination below 

 the ectoderm a single layer of loosely connected cells may be observed 

 extending into the cephalic lobes. A little more caudad in the region of 

 the stomodaeal invagination is the anterior cell mass several layers deep. 

 It is not separated or differentiated from the cells immediately posterior 

 to it. In the region of the mandibles and maxillae the mesoderm 

 extends as a single layer into the evaginations. The mandibles and 

 maxillae are now fully as long as they are broad. Two or three hours 

 later the anterior cell mass becomes somewhat reduced in size by cells 



