SKELETO-TROPHIO TISSUES AND COXAL GLANDS. 159 



same way as are the gland-caeca, so that it is not always possible 

 in a section to say which is a glandular caecum, and which is a 

 connective-tissue lacuna. (See the right hand side of fig. 1 in 

 Plate X, in illustration of this fact.) 



The fact also that there is no definite membranous capsule 

 to the coxal glands of Limulus, but that the lacunar tissue of the 

 prosomatic mass passes directly into the intercsecal tissue of 

 the gland, as sliown in Plate 10, fig. 1, at the line h h, renders it 

 not improbable that the whole gland — intercsecal tissue and csecal 

 glandular epithelium also — is a modification of the connective 

 tissue. It will be seen by Plate X, fig. 1, how difficult it is to 

 distinguish the lacunae e and the glandular caeca k. Possibly 

 this is a true suggestion as to the origin of the glandular 

 epithelium, both in the coxal glands of Limulus and in those 

 of Scorpio and Mygale ; possibly it is in all three a gland-cell 

 modification of the connective tissue. But until it is proved by 

 embryological research that this is the case, I shall adopt the 

 more probable hypothesis that the glandular epithelium is an 

 offspring of either the hypoblast or the epiblast, and that its 

 caecal growths are packed in and supported by mesoblastic 

 tissue — the intercaecal element in all three cases alike. 



The caecal epithelium in Limulus' coxal gland, when 

 favorably seen, presents distinctive characters. Its nuclei 

 are round (those of the intercaecal tissue are often oval) and it 

 has the same deeply placed, highly refringent, finely striated, 

 cortical layer which characterises the caecal epithelium of the 

 coxal glands of Scorpio and Mygale. The cells being much 

 smaller in Limulus than in the latter, the cortical layer is pro- 

 portionately narrow, but it is sufficiently obvious to give a 

 strong limiting line to the cell-layer on the side towards the 

 connective -tissue framework (PI. XII, fig. 4; cr). It is also 

 striated as in Scorpio and Mygale. 



The form of what I have called the " caeca " in Limulus for 

 the sake of comparison with Scorpio and Mygale differs very 

 much from the spaces so-called in these latter forms. The 

 substance of the coxal gland of Limulus is, in fact, honey- 

 combed by an irregular series of chambers — apparently all 



