228 SUMMABY OF OUBRENT BESEABCHES RELATING TO 



development we findasuperomeclian tooth with superolateral structures, 

 and in the other the tooth of the inferomedian joouch is well developed, 

 with its tooth-like lateral processes. The Alpheinse, Palsemoninse, 

 CrangoninsB, and Gnathophyllinse are in common characterized hy the 

 fact that the inferomedian and inferolateral regions of the cardiac and 

 pyloric portions are alone provided with hard structures of cha- 

 racteristic form ; the inferomedian cardiac process evidently consists 

 of three distinctly differentiated longitudinal portions ; of these the 

 median one has few or no setsB, while the lateral portions have short 

 setse, of various forms and arranged in groups, which look towards 

 the middle line; with this are placed longer setsB which form a 

 continuous fringe, and, when the gastric musculature is well deve- 

 loped, this fringe is provided with special muscles. The effect of 

 this arrangement is to confine the food-particles to the median line, 

 and to drive them along it into the thoracic region of the stomach. 

 Bearing this in mind, we can understand that the loss of the peristaltic 

 action of the stomach is due to the reduction of the cardiac process 

 and the great development of the pyloric superomedian process. 



The author concludes from his elaborate survey that the hard 

 structures of the stomach of the higher Crustacea are most im- 

 portant aids in the classification of these forms ; and his own results 

 coincide with those arrived at by v. Boas. The Natantia form 

 the lowest groups, and the Eucyphotes may be defined as Decapods 

 without a cardiac dorsal mill, and the Penasidae as Decapods provided 

 with one. The Atyinae appear at present to be isolated forms, but a 

 connecting link may perhaps be found in Troglocaris, The Serges- 

 tidfe are to be placed with the Penseidse, as are also the Gerataspis 

 forms, which are often associated with Schizopoda. The well-defined 

 .group of the Homaridse may be divided into the Homaringe and 

 AstaciuEB, as Boas has suggested. The Anomala (in the sense of 

 De Haan) do not form a definitely separated group. 



The type of gastric mill found in the Decapoda may be continued 

 into the Squillidee, Mysidse, and Cumacese, where almost all the 

 corresponding parts are to be found. The median inferomedian pyloric 

 process forms a crest-like longitudinal invagination, and passes 

 through very interesting gradations ; in Diastylis it has one, in 

 Mysis two, in Gammarus three, and in the higher Malacostraca a number 

 of fringes of longitudinally-set setee ; indeed, the number increases as 

 the form possessing them stands higher in a systematic classification. 



Spermatogenesis in Hedriophthalmate Crustacea.* — G. Herr- 

 mann finds that the spermatogenesis of hedriophthalmate is effected 

 on a different plan to that of the podophthalmate Crustacea. The 

 male cells are of large size, and, soon, come to have a number of 

 nucleoli, around which the nucleus seems to undergo segmentation. 

 This is followed by a stage in which there is a group of smaller 

 nuclei, irregular in form, more or less definitely arranged round the 

 periphery of the sperm-cell (" ovule "). This latter divides into equal 

 parts, and soon small cell-elements, with a nucleus, but without a 



* Comples Kendus, xcvii. (1883) pp. 1009-12. 



