1909.] NATURAL SCIENCES OF PHILADELPHIA. 307 



wall, resembling to some extent the primary folds of the pulmonary 

 invagination. After the stage of reversion he figures it as a simple 

 tube without branching. The most recent paper is that of Janeck 

 (1909); he did not see the earliest stages, but describes and figures a 

 stage shortly after hatching with the beginning of the branching, and 

 an advanced stage after reversion. 



Accordingly, no one has figured the earliest stages of the trachea 3 , 

 and our knowledge of their development is very fragmentary. 



In Loxosceles the conditions are as follows, as observed in surface 

 views only. In the stages of figs. 17-19, Plate XIII, the appendages 

 of the third abdominal segment (3) are simple rounded protuberances 

 like those of the adjacent segments. Later, fig. 20, the appendage of 

 the third segment becomes flatter and less distinct, and its elevated por- 

 tion has taken on somewhat the shape of a quarter moon. At the 

 stage of fig. 21, completed reversion, no elevated appendages are found 

 in that segment, but at its posterior edge a transverse groove circularly 

 enlarged at each lateral corner (T. St.). This condition persists 

 through the stages of figs. 22 and 23, but at that of fig. 24 a pair of 

 lateral tracheal branches, recognizable by being filled with air, and 

 marked Tr., are growing forward from the groove. The transverse 

 groove is evidently the stigma leading into the vestibulum; and this 

 stigma, developing at the posterior edge of the third segment, evi- 

 dently arises behind the appendages of the segment. In the adult 

 a larger median and four pairs of smaller lateral tracheae extend for- 

 ward from the vestibulum (fig. 27). 



A better knowledge can be obtained from the study of sections 

 which I have made of Theridium. At the time of reversion (fig. 16, 

 PL XII) the abdominal appendages show much the same relation as in 

 Loxosceles, though a narrow elevated ectoblastic ridge joins the second 

 and third abdominal appendages (£. Ap., 3. Ap.). Sections of this 

 stage (figs. 1, 2, PL XI) show the still heightened appendage (3. Ap.) 

 of the third segment behind the pulmonary invagination and a distinct 

 coelomic sac (Coel.) appertaining to it. But in subsequent stages 

 (figs. 5, 7, 9) the appendage of the third segment disappears, the 

 whole becoming flush with the remainder of the surface of that seg- 

 ment and without any process of invagination. It is not until after 

 this appendage has disappeared entirely that the tracheal invagination 

 takes place, and evidently as in Loxosceles behind the region where 

 that appendage was. The first appearance of the tracheae is seen in fig. 

 12, PL XII ; the pulmonary stigma (P. St.) is at the posterior edge of the 

 second segment, the tracheal (T. St.) not near this, as was the append- 



