1909. 1 p. Speiser : The Family Polyctenidce. 273 



wanting. The thorax has an outline somewhat different from that 

 of the imago ; it is widest a Httle before the middle of its length, 

 whilst it reaches its greatest width in the imago behind this point. 

 The arrangement of the bristles is exactly the same as in the 

 imago, but the oblique rows of longer bristles on the anterior half 

 contain on ea^h side five instead of thrse bristles. The concave 

 round impressions on the lateral and hinder parts are absent ; 

 they are wanting also on the elytra, which are a little shorter than 

 in the imago. Their length is not greater than four-fifths of the 

 width in the anterior part. The arrangement of the bristles on the 

 abdomen is closely similar to that of P. talpa, mihi, as figured by 

 me (1904) in the description of this species. The first segments are 

 in the nymph also naked on the disk and bear on the hind border 

 a simple row of very short bristles. But they bear at the posterior 

 angles of the tergites on each side one longer bristle, which is not 

 present in the imago. In this latter the segments vi to viii of the 

 tergum are beset on their surface with many fine bristles, and on 

 the hind border with considerabl}' longer and stronger ones ; 

 while in the nymph they are all quite bare on their disk, and the 

 seventh and eighth only bear the longer marginal bristles. The 

 bristles on the sternal surface are also thinner than in the imago. 

 The sternites bear scarcely more than a single row of very fine 

 bristles, and in some cases in the middle before this a small group 

 of similar ones. 



A matter of peculiar importance is the structure of the tarsi. 

 In the Polyctenes as yet known they must be described as four- 

 jointed, but in this nymph they are clearly three-jointed in both the 

 hinder pairs of feet. By this observation I find confirmed, in a very 

 gratifjdng manner, what I asserted in 1904, namely, that the second 

 and third tarsal joint of the Polyctenidae is homologous with the 

 second or middle tarsal joint of the other Hemiptera. For in the 

 nymph now known this joint is, in fact, as I postulated then, the 

 second tarsal joint not yet divided into two. From this statement 

 arises a new and unexpected support to my views on the system- 

 atic position of the family Polyctenidce. 



The important results reached by the finding of this new 

 specimen of the rare family may be enumerated as follows : — 



( 1 ) The Pol yctenidse are now with greater accuracy than before 



put in close affinity with the Acanthiadae (Cimicidae). 



(2) The ctenidia on the head are phylogenetically older 



characters than those on the elytra, and the latter 

 are older than those on the thorax : they appear 

 within the individ lal Hfe in this succession: the 

 ctenidium on the elytra appears at a time when a 

 ctenidium on the thorax is lacking (in P. lyrce, this 

 being a nymphal stage). 

 (3)^- spasmcs, Waterh., may be perhaps a larval stage of 

 P. talpa, mihi; should this be proved by specimens 

 found later, the much older name given by Water- 

 house has the priority. 



