MESODERMAL DERIVATIVES 117 



sinus. Philiptschenko (1912) failed to find an epineural sinus in the 

 collembolan Isotoma, but it may have been overlooked. 



CIRCULATORY SYSTEM 



The development of the dorsal blood vessel, or heart, and of the dorsal 

 diaphragm, except for minor differences, is similar for all the Eutracheata. 

 The origin of these structures was first observed by Korotneff (1883, 1885) 

 in the mole cricket {Gryllotal'pa vulgaris), and since then the further 

 investigations of Grassi (1884), Heider (1889), Wheeler (1889), Heymons 

 (1895a), and others working with various insects have fully confirmed his 

 findings. The cells that are to form the walls of the heart, known as 

 "cardioblasts," are differentiated at the dorsal junction of the somatic 

 and splanchnic mesoderm layers. As these layers grow dorsad, the 

 cardioblasts, which first are recognizable by reason of their larger size, 

 become, in certain orders at least, more or less crescent-shaped, their 

 convex sides directed toward the ectoderm, the concave sides of the two 

 rows facing each other (Fig. 21). When the horns of the crescents 

 finally meet, a longitudinal, middorsal tube is formed, constituting the 

 heart. The closure of the heart tube, in general, first takes place at the 

 posterior end of the embryo, later in the neck region where the heart 

 connects w^th the aorta. In Donacia Hirschler (1909) noted that the 

 fusion of the cardioblasts first occurred on the ventral side at the posterior 

 extremity of the embryo. At the anterior end in the region of the 

 stomadaeum, on the other hand, the dorsal wall of the heart is completed 

 first. In the genera Donacia, Chrysomela, and Formica, the cardioblasts 

 are rounded and not crescentic at first. With some insects, before the two 

 rows of cardioblasts meet, between the cardioblasts and the yolk there 

 is a blood space filled with fluid. This space, especially large in the 

 crickets and some acridids, is less well developed in Forficula, in Coleop- 

 tera, and in the roaches. In some cases pulsation has been observed in 

 these blood sinuses. In Locusta migratoria Roonwal (1937) records the 

 formation of a primary pair of lateral blood sinuses which are bounded 

 dorsally by the provisional dorsal closure and ventrally by the somatic 

 mesoderm and sometimes also by the ectoderm forming the lateral edges 

 of the germ band. During blastokinesis, the provisional dorsal closure 

 (ental membrane) breaks away from the edges of the germ band and 

 curls upward to bound the yolk for some time. The amnion, which 

 now forms the dorsal closure of the embryo, becomes joined to the 

 provisional dorsal closure middorsally. Thus a second pair of lateral 

 sinuses arises in place of the first one. They are bounded dorsally by the 

 amnion, ventrally by the provisional dorsal closure (afterward by the 

 splanchnic mesoderm), and laterally by the somatic mesoderm. The two 

 lateral sinuses later fuse to form the common dorsal blood sinus, and 



