Chilton. — On New Zealand Freshivater Crayfish. 245 



for easy reference, are purely arbitrary, and may not always 

 be clearly distinguished. The spines often vary both in num- 

 ber and position on the two sides of the same specimen. In 

 some specimens the spines are almost entirely absent, so that 

 the whole of the carapace is nearly smooth ; but usually some 

 of the groups are represented, if not by spines, by small 

 tubercles corresponding in position to the spines in other 

 specimens. I have tabulated the number of spines of each 

 group found in various specimens, and, though I need not give 

 the tables, I can summarise some of the results here. Group A 

 contains only two spines, and is invariably present ; group B is 

 frequently altogether absent, but ma.y contain several spines, 

 as in the Nelson specimens, extending into group D and group 

 K, mentioned below ; group C is usually represented, but often 

 only by very small spines ; group D is often absent in small 

 specimens, but may contain as many as six spines, extending 

 right back to the cervical gi'oove ; group E is seldom entirely 

 absent, and never contains more than two spines ; group F is 

 ofcen absent, and the spines in it are never large ; group G is 

 one of the most constant, and contains some of the largest 

 spines, it is seldom entirely absent ; group H, again, is also 

 usually represented, though the number of spines varies very 

 much, and they are sometimes very small and close together, 

 so as to form only a row of more or less spinous tubercles ; 

 group J, including the spines and tubercles on the branchio- 

 stegites, is very variable — sometimes the spines are entirely 

 absent, so that this part of the carapace is quite smooth, at 

 other times the spines are represented by slight roughness 

 only, but, again, as in the Nelson specimen, the whole of the 

 branchiostegite may be thickly covered with well-developed 

 spines. In the specimen from Nelson, two other gi-oups, pre- 

 sent in many specimens of P. nco-zelanicus, but not represented 

 in other specimens of P. lAaivifrons , are noticeable : they are, 

 gi'oup K, containing 4-5 spines in front of the cervical gi'OOve 

 and behind gi'oup B; and gi'oup L, containing two or three 

 small spines on the cardiac portion of the carapace, usually by 

 the sides of the branchio-cardiac grooves. (See Plate X. 

 fig. 2rt.) In this Nelson specimen the number of spines in each 

 group is much greater than usual, and they run into one 

 another so much that, as in some specimens of P. neo-zelanicus, 

 it is difficult to distinguish them into the different groups. In 

 the Lake Eoto-iti specimens the spines are usually more or 

 less blunt, and often degenerate into tubercles. 



The squame of the antenna sometimes reaches only as far 

 as the extremity of the rostrum, instead of beyond it as in the 

 Napier specimen. The inner edge is geiaerally straight in the 

 centre, or even slightly concave ; but sometimes more convex, 

 as it usually is in P. nco-zelanicus. Towards the end it often 



