136 THE AMERICAN XATURALIST. [Vol. XXXVIII. 



scope the dorsal surface is seen to be divided into polygonal 

 areas. 



The appearance of the cells of the dorsal layer varies much 

 with the treatment of the tissue. Material fixed in alcohol and 

 brought in contact with water for but a short time in staining 

 with hematoxylin shows the cells filled with a dense content of 

 coarse granules. With reflected light these granules still show 

 their characteristic chalky whiteness, while with transmitted 

 light they are brown. Granules identical in appearance are also 

 found in the fat cells of the same region of the body. Material 

 fixed in any fluid requiruig subsequent washing in water shows 

 a considerable decrease in the 

 granules of the dorsal layer. 

 This verifies the statement of 

 Wielowiejski that these gran- 

 ules are insoluble in alcohol, 

 but soluble in water. Kolli- 

 ker, in 1857, proved them to 

 be crystals of urate salts, and 

 his results have been accepted 

 by Wielowiejski and Emery. 

 W^hen the crystals have been 

 dissolved out the form of the 

 cells is easily determined. 

 They are polygonal, fairly 

 regular in outline and simi- 

 8 by 25 /X. Those upon the 

 spherical than those beneath. 

 Large nuclei are always present, but the cytoplasm seems to 

 have been almost entirely replaced by the granular secretion. 

 (Fig- 3-) 



The dorsal layer not only forms a plate resting upon the 

 ventral layer, but it projects beyond the latter and extends along 

 its caudal surface to the body wall (Fig. 2). There are two 

 groups of muscle fibres in each light plate, extending from the 

 dorsal to the ventral body wall in the lateral portions of the 

 plate. These muscles are surrounded by a layer of cells dis- 

 tinctly separated from the cells of the ventral layer and contin- 



