REPORT ON THE RADIOLARIA. 7(37 



spiues also not clifFermg from one another (Acanthostauriis, &c.). But in the majority 

 of this family there are three ditiereut kinds : four larger equatorial spines, eight tropical 

 spines of middle size, and eight smaller jjolar spines (Belonostatmis, &c.) ; the latter 

 become sometimes rudimentary, so that only twelve spines are developed (four larger 

 equatorial and eight smaller tropical spines). In this case the development of the flat 

 discoidal body is much stronger in the equatorial plane than in all other planes. 

 These discoidal or lenticular Quadrilonchida exhibit a relation to the spherical Astro- 

 lonchida similar to that which the Discoidea exhibit to the Spheeroidea among 

 the S p h iB r e 1 1 a r i a. 



A further morphological differentation takes place in the remarkable genera 

 Lonchostaurus and Zygostaurus. In the former (PL 131, figs. 4—6) the four larger 

 ecj^uatorial spines becomes differentiated in pairs, so that the opposite equal spiues of 

 one pair, cl, c3 (in the longitudinal or hydrotomical axis), are larger, and often also of 

 another form, than the opposite equal spines of the other pair, c2, c4 (in the transverse 

 or geotomical axis). The most peculiar form is the rather common Zygostaurus 

 (PI. 131, figs. 7, 8). Here the two opposite spines of one ec|uatorial axis (of the 

 longitudinal axis) become very different, so that the anterior or frontal spine (cl) is 

 very unequal to the posterior or caudal spine (c3), whereas the two opposite sjiines, 

 of the other equatorial axis (of the transverse axis) remain equal (c2, c4). 

 Therefore the fundamental forms become here " amphithect," as in the Ctenojjhora. 



The numerous Quadrilonchida may be disposed in two different subfamilies : in the 

 Acanthostaurida all twenty radial spines are sinqjle, without apophyses ; in the 

 Lithopterida all twenty spines (or only one part of them) bear two opposite apojihyses 

 (or lateral transverse processes). The former correspond to the Zygacanthida, the 

 latter to the Phractacanthida among the Astrolonchida. The two opposite apophyses 

 are simple in Quadrilonchc (PI. 133, fig. 1). In Xiplioptera they are provided on 

 their distal side with lateral branches which are parallel to the spine itself. In 

 Lithoptera the spines bear two to four parallel pairs of transverse apophyses, and 

 these are crossed by perpendicular branches, parallel to the spine itself, so that there 

 arise fenestrated wings or latticed plates, comparable to the sails of a wind-mill. 

 The lattice-work of these plates lies in the same meridian plane with the radial spine 

 itself, and is therefore not comparable to the fenestrated apophj^ses of Doracantha, of 

 Phatnacantha, and of the Dorataspida ; in these the lattice-plates lie in tangential 

 planes, perpendicular to the radial spine. 



The apophyses of the Lithopterida may be developed either on all twenty spines 

 equally, or only on twelve spines (four equatorial and eight tropical, whilst the eight 

 polar spiues are simple, PI. 131, fig. 10), or only on the four equatorial spines (whilst 

 the sixteen others are simple, PI. 131, fig. 9). 



T7ie Central Capsule of the Quadrilonchida is rarely spherical, commonly more or 



