1909.] NATURAL SCIENCES OF PHILADELPHIA. 315 



Simmons insisted on the following points: (1) that primary folds 

 appear very early on the posterior surface of the invaginating pul- 

 monary appendage of the spider, resembling very closely in position 

 the gill-leaf development of Limulus as described by Kingsley ; and 

 (2) that such folds become the pulmonary lamellae of the adult spider. 

 But Janeck and I demonstrate that these primary lamellae of the spider 

 are only temporary, with probably no more significance than the folds 

 on the mesial surfaces of the cephalothoracal appendages, and they 

 certainly do not become the definitive lamellae of the lung-book. The 

 truth is that the secondary pulmonary plates of the spider do not 

 arise from folds on the outer surface of an appendage, but in a differ- 

 ent position from a solid cell mass after the appendage has invagi- 

 nated. 



Therefore there is nothing in the ontogeny of the lung-books of spiders 

 similar to the development of the gills of the Xiphosura, beyond that 

 they are both connected with abdominal appendages; consequently 

 no embryological ground for deriving the lung-books from gills. 



Further, objections of a theoretical nature may also be made against 

 the derivation of the lung-books from gills. First, the direct change 

 of a water-breathing organ into an air-breathing one would be almost 

 without parallel; at least I cannot recall any instance of this kind in 

 animals. Second, granted that Limulus is genetically related with 

 arachnids, a view to which I subscribe, what reason have we for 

 assuming that Limulus is the more primitive form? The only good 

 anatomical argument for this view is that Limulus retains through life 

 six pairs of abdominal appendages, and also a pair of nephridia. The 

 general reason for the primitiveness of Limulus is its marine habitude, 

 with the assumption that the ancestors of most groups must have 

 originated in the seas. The latter assumption lacks all solid basis, 

 and has been opposed by Simroth (1891) and myself (1906). But the 

 true arachnids are essentially terrestrial, quite as much as the insects 

 are, and though a few have taken to aquatic existence, such as the 

 Hydrachnids and Argyroneta, all of them breathe atmospheric air. 

 Further, true araneads occurred in the Carboniferous, in times as early 

 as the first fossil traces of Limuloids, while terrestrial scorpions existed 

 in the Silurian. These facts would indicate that the true arachnids 

 originated on the land. It would, accordingly, seem to me that if 

 Limulus be related to the arachnid stock, and this seems well estab- 

 lished, that the former represents a line of descent that has taken 

 secondarily to aquatic life ; and for this view would speak also its exist- 

 ence in shoal waters, and its behavior in laying its eggs close to the 



