I28 NEW YORK STATE MUSEUM 



many crustaceans, " while the ordinary compound eyes on the free 

 cheeks are absent." In their structure these simple eyes of Harpes 

 and Trinucleus may be called ocelli, but they are not homologous 

 to the median or parietal eyes of the crustaceans, as the last clause 

 of Beecher's sentence would indicate, and we consider them as 

 either primitive or atrophied, lateral, compound eyes, agreeing in 

 this regard with Clarke (no. 4) who unites them as simple, par- 

 ticular cases with the schizochroal eyes. Our reasons for this 

 view are : 



1 The " ocelli " of Harpes and Trinucleus are situated at the ends 

 of the primitive " eye lines " which, according to Beecher, are 

 found in four-fifths of the Cambric trilobites. But these Cambric 

 trilobites show at the end of the eye lines regular compound lateral 

 eyes (as Olenus, for instance) situated along sutures. Hence the 

 " ocelli " of Harpes and of the young Trinucleus are entirely homo- 

 logous to the compound eyes of the other trilobites, as far as 

 regards their terminal position to the eye lines. 



2 The " ocelli " of Harpes show as many as three lenses each, 

 while the median eyes of the crustaceans show no more than two 

 lenses where they are united in the middle of the carapace, or one 

 ocellus for each eye. 



3 The " ocelli " of Cryptolithus (Trinucleus) at least can not be 

 homologous to the median eye of the crustaceans because we have 

 found in that genus the true median eye on the glabella (see below), 

 and the facial suture connected with the so-called " ocelli " (see 

 p. 144). 



The median eyes which are in this note claimed for the trilobites, 

 appear in the majority of cases as a single tubercle upon the gla- 

 bella, mostly on a line connecting the posterior extremities of the 

 lateral or compound eyes. We first observed these minute tuber- 

 cles on the glabellas of half or less mature specimens of I s o - 

 telus gigas (see pi. 34, fig. 6). While the entire integu- 

 ment of this trilobite is smooth, that is, only finely punctate but 

 entirely without any tubercles, there is always seen this single fine 

 tubercle on the posterior portion of the glabella. It is most distinct 

 on the earliest growth-stages and mentioned and well figured in 

 Raymond's careful " Notes on the Ontogeny of Isotelus 

 gigas Dekay " (7, p. 250, fig. 1; pi. 7, figs. 1-3) as a "median 

 tubercle."' We then found this tubercle to be a constant feature 

 in many genera (about thirty, see below under 8, p. 138) where 

 as such it had already been recognized by various authors for one 

 or several species in each case, thus by Barrande (no. 1) and by 

 Oehlert (no. 12), for all species of Cryptolithus (Trinucleus), and 



