SCIENCE- G OS SIP. 



!55 



the papillae of the lateral organ. The halteres 

 receive their rich supply of nerves direct from the 

 second thoracic ganglia. This pair of nerves is 

 the largest in the thorax, and crosses to the opposite 

 side immediate!}' on entering the ganglia. The 

 pedicle is a hollow tube connecting the base of the 

 halter with the globe. On the external surface 

 it is covered with hairs. The interior is divided 

 by a septum which is continued the whole length. 

 A large tracheal vessel passes through it to the 

 globe, where it breaks up into many branches 

 which ramify in the tissue. 



Sarcophaga Carnaria, Longitudinal Section, of 

 Halter. — These sections show the vascular tissue 

 in the so-called globe of the halter. In all the 

 halteres I have examined the deep invagination seen 

 in these sections of the globe is invariably present, 

 and there is always a mass of connective tissue 

 extending from the invaginated wall to the oppo- 

 site wall of the globe. The purpose of the in- 

 vagination is unknown to me, unless by some 

 means it allows of a certain amount of expansion 

 and contraction of the globe. The large glands 

 most probably secrete a fluid necessary for organ* 

 at the base of the halter. The halteres of Diptera 

 doubtless assist in their locomotion, but the evi- 

 dence of their elaborate structure proves that they 

 have another most important function. The posi- 

 tions of the papillae are such as to present a front 

 in every direction, and their structure is so delicate 

 as to permit of vibration when sound-waves or 

 other unusual movements of the air impinge upon 

 them. Also the nerve epithelium bathed in fluid 

 secreted in the globe, together with the very rich 

 nerve supply, point to their being rudimentary 

 nerve organs. Otoliths, so commonly found in the 

 Crustacea and Mollusca, I have not met with here, 

 but that does not prove their non-existence. The 

 great number of papillae (400 to 500) in each 

 halter, and the small number of olfactory organs 

 (two in each antenna) found in many flies which 

 feed on the nectar of flowers, compared with 

 M. vomitoria and M. domestica, whose halteres 

 carry half the number of papillae, and in whom 

 the olfactory sense is highly developed, show that 

 the former possess an acute sense to warn them of 

 danger when their heads are buried in the blossoms 

 of the plants they frequent, and that the latter 

 have comparatively little use for such a sense. 



Anterior Thoracic Spiracle of Blow-fly (fig. 5, ante, 

 p. 123). — This spiracle is oval and narrowest above, 

 and is situated between the pro- and meso-thorax. 

 From the exterior free edge project hollow arbor- 

 escent chitinous rods, which curve outwards and 

 interlock for about one-third of the length of 

 the spiracle. These rods are hollow, even to the 

 minutest twigs, which have a free opening at their 

 points. Close behind is a transparent membrane, 

 the true valve. It is united to the wall of the 

 large tracheal vessel which extends across the 

 thorax to the opposite spiracle. The free edge of 

 the valve is closely set with a chitinous fringe. A 

 special muscle arises from the integument at the 

 lower end of the spiracle. By the contraction of 

 this muscle the free edges of the valve would be 

 caused to approach each other. From the integu- 

 ment another set of muscles arises, which are 

 directed towards the valve, but whether they are 

 connected with it I have not been able to deter- 

 mine. Antagonistic muscles are a necessary con- 

 sequence for working the valve. 



Posterior Thoracic Spiracle of Bloic-fly (fig. 6, ante, 

 p. 123). — In structure this spiracle is very similar 

 to the anterior one, excepting that the external 

 chitinous rods are formed into two distinct masses 

 by the addition of a connecting membrane. 



Posterior Thoracic Spiracle of Sericomyia Borea- 

 lis (fig. 7). — This spiracle differs from the above in. 



Fig. 7. Posterior thoracic spiracle of Sericomyia borealis. 



having the chitinous rods free, and also behind* 

 the valve the walls are lined with folds of mem- 

 brane with the edge directed inward. 



Abdominal Spiracle of Bloov-fly (fig. 8). — The 

 abdominal spiracles are very much smaller than 

 those of the thorax. Their relative sizes may be 

 seen in the illustrations, which are of the same 

 magnification. The spiracle is round, fringed 

 with fine hairs, and the valve is placed a very 



FlG. 8. Abdomiuil spiracle of Musca vomitoria. 



short distance behind it. The valve appears to^ 

 consist of a thickened membrane on the one side r 

 which gradually thins out towards the free edge. 

 The other half is thinner and more flexible, and its- 

 movement is effected by a curved rod or bow, 

 hinged at one end. The other end is connected 

 with a set of muscles arising from the edge 

 of the spiracle. No antagonistic muscles have 

 been found ; but if air is both received and 

 expelled by it, such muscles are requisite. It 

 is the generally-accepted theory that all the spi- 

 racles are both afferent and efferent ; but, judging - 

 from the observations of others as well as my- 

 self, I must confess that I am somewhat doubtful 

 on the point. The minuteness of the opening of 

 the abdominal spiracles, and the almost immediate 

 branching of the large tracheal vessels are eminently 

 suited to exclude dust particles, and for the quick 

 distribution of inflowing air ; added to these, the 



