MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



423 



can be said. Brooks tliought that in Lucifer they represented the food yolk, although this was 

 not settled. In Alplieus these cells are later joined bygrer.t nnnihers of wandering cells at the 

 invagination period, and out of this common stock, so far as we can detormiiie, both raesoblastic 

 and eutoblastic organs are developed. In Horaarus, on the other hand, the invaginate cells 

 unquestionably degenerate. 



The eggs of Scullarus and PaUmtruit have not been thoroughly studied, so far as I am aware. 

 Dorhn (14), however, has figured the eggs of Scyllarus arctus in late stages of yolk segmentation, 

 and in surface view they resemble, at this time, the ordinary Decapod type. We have as yet no 

 knowledge of the .segmentation of the Storaatopods. 



The relations of the diflerent varieties of segmentation which are met with among the Deca- 

 pods, may be expressed by the following table : 



(I). Lucifer typus. 



Segmentation of the . 

 Decapod egg. 



I. Total: regnlar: seg- 

 mentatiou cavity pres- 

 eDt. 



II. Partial : centroleci- 

 thal(yoIk pyramids): 

 Begmentaticiii cavity 

 filled with yulk. 



Segmentation of yolk at first to- 

 tal, afterwards partial. 



^(3 



(•Regular. < 



(II). Segmentation of 

 yolk. { 



Irregular 



(1) ralcmion. 



) Eupagnrus priilcauii. 

 (3) Atyephi/ra compressa. 



(1) Penaus. 



(2) Crangon. 



(3) Stetiopun hispidiis. 



(4) AlpliciiK naulcyi. 



(5) Ponlotiin domcsiica. 

 (fi) Hippa talpoidex. 



(7) ral(vmo»(lrn I'ulgaris. 

 , (8) Ciillianaesa mcditerranea. 



(1) HomaruK americanux (yolk 



segmentation at first ir- 

 regular, but later regular 

 or nearly so). 



(2) Alphens minor. 



Of the Thoracostraca, the Schizopods undoubtedly depart widest from the common decapodal 

 type of segmentation. Nusbaum (45) thus describes the process in Mysis Chamrlco : The pro- 

 toplasm — that is, the segmentation nucleus, with its protoplasmic body — retreats to a point at the 

 surface of the egg. The nucleus then suflers division, and the protoplasm becomes diiferentiated 

 into an outer striated zone containing a single nucleus, and an inner granular zone with one or more 

 nuclei. The single e.xternal necleus divides and gives rise to a small blastodermic di.sk, formed of 

 a single layer of hexagonal cells. From the internal nuclei and protopla.sm a small number of nuclei 

 are produced under the blastodermic disk. The free cells below the disk are the products (1) of 

 the nuclei and formative protoplasm of the deeper layer, and (2) of the cells of the upper layer or 

 blastodermic disk. While a solid accumulation of cells is thus being formed below the disk, the 

 superficial cells gradually extend on all sides and inclose the egg. The thickened disk marks the 

 ventral side of the embryo. It divides into a median, caudal, or abdominal plate, and two lateral 

 plates, the rentral hands. Volk cells which were not present up to this stage, now arise by migra- 

 tion from the abdominal plate. 



Tlie segmentation of the Schizopod is especially interesting, since it agrees so closely with that 

 described in certain Isopods and Myriapods, and resembles also the segmentation of Arachnids. 



Bobretyky's observations on Oniscus {'>) need to be repeated, and especially in the early 

 stages. Acconling to this observer the earliest phase of segmentation was characterized by the 

 anomalous withdrawal of the egg protoplasm to the surface, where it accumulated in a distinct 

 body, and underwent segmentation. A di-sk of large columnar cells was thus formed, marking the 

 ventral surface of the embryo. The cells spread with the thickening of the disk, until they inclose 

 the entire egg. The superficial cells form ectoblast, the rest entoblast and mesoblast. 



Morgan (41) has de8cril>ed a form of .segmentation in Pycnogonida which resembles that of the 

 Decapod. Here the invagination (which leads to Uie formation of the stomoda'um) is preceded by 

 the delamination of endoderm cells from the blastospbere, very much as in Alpheus, if we may 

 regard the yolk cells a.s prioMtive endoderm. 



