18 LOWER INVERTEBRATES. 



name to the nummulitic limestone, the tubes have a different arrangement and are 

 very minute ; there is, besides the tubular structure already described, a system of 

 inosculating canals penetrating the septa, which are filled with sarcode during the 

 life of the animal. In all the vitreous forms, each chamber has its own shelly in- 

 vestment, so that the partitions between the chambers are double. Between these 

 walls there is frequently a considerable deposit of calcareous substance, which is 



known as the intermediate skeleton. Through the 

 intermediate skeleton runs the system of canals, 

 which is beautifully shown in Eozoon, soon to be 

 described; the canal-system resembling minute 

 branching shrubs. A single species of Numnvu- 

 FiG. \^. — KummvMim wiicoxi, natural size, Una, Fig. 16, has been described from Florida. 



and enlarged. t-. -i i • . «. i n i 



Besides the two varieties oi shell-structure above 

 described, there is another kind of shell or test very frequently occurring among deep- 

 water sjDecies. This is the arenaceous type, in which the shell consists of cemented 

 grains of sand, or of sand and spicules together. 



The nature of the cement which holds together the sand-grains of the arenaceous 

 types is not known ; sometimes the grains are only loosely united, so that the test is 

 more or less flexible, as in Astrorhiza, Fig. 17, a form which is found in Vineyard 

 Sound at depths of only twelve fathoms, but also reaching down to over five hundred 

 fathoms. Some of these have the outside test smoothly plastered by a layer of very 

 fine particles of mud, although composed of irregular large and small grains of sand. 

 No definite aperture, or mouth, has been observed in Astrorhiza, 

 and the sarcode finds its way through the test between the loosely , 



cemented grains composing it. In other forms the grains are 

 very closely cemented, so that some tests will resist the action 

 of warm nitric acid, proving that the cement is neither calcareous 

 or ferruginous. In some cases the sand-grains seem to have a 

 chitinous basis in which they are imbedded. The resemblance 

 between the arenaceous Foraminifera and the porcellanous and 

 vitreous species is striking. Take, for example, Salophrag- '^ 

 mium, and compare it with Globigerina, Fig. 18. "^ 



Indeed, it is true that if we consider only the external forms, largedTwo'dSneters."" 



we can find in the three divisions of porcellanous, vitreous, and 



arenaceous forms many species that are so closely related as to be indistinguishable 

 by any specific characteristics. Thus, Cornuspira among the porcellanous is the 

 counter-part of S^^irilUna among the vitreous forms ; and this is distinguishable in 

 form from Ammodiscus among those with sandy tests. 



While some of the tests of the arenaceous gi'oup are probably imperforate, others 

 are, without doubt, more or less porous, so that the distinction already made between 

 hyaline and porcellanous forms must also hold good as concerns these. Indeed, cer- 

 tain arenaceous forms have no definite mouth, and the sarcode must find ite way 

 through pores in the test. 



The deep-sea investigations that have been carried on of late years have brought to 

 light many new foi-ms belonging to genera which were supposed to be very well known. 

 Thus, the shell of Globigerina, Fig. 18, has been understood, conforming to the descrip- 

 tion of Dr. Carpenter, to consist of a series of hyaline, perforated, spheroidal chambers 

 arranged in a spiral about an axis, each opening into a central space in such a manner 



