Respiration in Invertebrate Animals. 307 



its elements anew. This change continues after the fluid escapes 

 from the cells, beyond the pale of the living solids. The oil-cells 

 (c) in the axis of the follicle are not organized cells, as represented 

 by Dr. Leidy. They are mechanically-formed oil- and albumen- 

 cells {g) . The chemical change proceeds in the secreted product, 

 icithout the direct agency of cells, as it flows from the distal to 

 the proximal end of the tubule*. The yellow colour {h, d) is not 

 developed until the product reaches nearly the middle of the 

 tubule. The process of respiration can only be resolved to its 

 last analysis by first determining with exactness the manner in 

 which other vital nutritive acts are accomplished. The endosmose 

 of gases requires that the partition should be thin (PI. XYIII. 

 fig. 5, a, b), the epithelial cells compressed and attenuated, and all 

 unnecessary contents withdrawn ; the act of fluid secretion de- 

 mands the presence of the reverse conditions (fig. 10, b) : such con- 

 trasted reasoning must prove directive to the thoughtful reader. 

 Induction founded on contrasts serves often best to establish the 

 principia of a science. 



The branchiae of the Anomourous Crustacea are arranged iii the 

 thoracic chamber in alternate series of fourteen in number. The 

 chamber is open along the whole extent of its inferior boundary. 

 To this circumstance is to be asci'ibed the absence in these 

 Crustaceans of the flabella. They are not required. The influx 

 and efflux of the water into the respiratory cavity are without 

 such assistance unimpededly rapid. Propelling and guiding 

 instruments would here prove a useless incumbrance. The gills 

 in the Paguridse are composed of laminae. They are four- 

 sided pyramids in shape. The respiratorj'^ leaflets are aiTanged 

 however in a bipyramidal manner (PI. XVII. fig. 7, B) ; that 

 is, a wide channel (between / and g, fig. 7, B), coinciding with 

 the length of the gill and with the primary blood-vessels, lies 

 between the rows of leaflets wbich are built up on each side of 

 the axis. Along this groove (fig. 9, a, fig. 7, d) rushes a cm-rent 

 of water, descending along the gill on one side and ascending on 

 the other. The ascent occurs on the inner side. The minutely- 

 divided streams of water, which directly operate on the respi- 

 ratory laminse (or cylinders, as the case may be), connect these 

 two main vertical currents by horizontally passing between the 

 branchial leaves. If these laminae were to become adherent 

 through the absence of moisture, the respiratory process could no 



* I have not attempted in the text, because it would be out of place, to 

 compare the results of my own investigations on the structure of the liver- 

 follicle of the Crustacea with those of Mr. Goodsir and Dr. Leidy. Such 

 comparison must be made by the future student. I venture to think that 

 neither the structure of the hepatic follicle nor the process of secretion has 

 ever before been placed in so clear a light. 



20* 



