BY R. J. TILLYARD. 217 



in the forewing, ;ij.':aiiist this claim. The proper degree of relationship between 

 the Thyrididae and the Butterflies is probablj best expressed if we say that the 

 old Protocossid stem gave origin in ascending order to (a) the Zygaenoid group 

 of families (Meyrick's superfamily Psyt-liina), (b) the Pyraloid group of fami- 

 lies, of which the most archaic, having very distinct affinities with the Tortneina. 

 is the Thyrididae, though these are not even to be considered as directly ancestral 

 to the Pyralididae themselves, and (c) the Rhopalocera, as a very distinct series 

 whose origin is to-day lost to us. 



At this stage we must leave this interesting problem, until such time as 

 pupae of the Zijgaenidae and Pyralididae can be obtained and their wings 

 studied with a \-iew to ol)taining further light upon it. Sufficient evidence has, 

 however been obtained to show the extreme unlikeliness of any existing Hetero- 

 neurous type representing e\en a close apjiroximation to the original archetype 

 of the Rhopalocera 



Corrigenda to Part 3. 



p. 5(31, line 14 from bottom, for "one'' read "two." 



p. 588, for "Text-fig. 53" read "Text-flg. 58." 



p. 594, remove the Lepiduptera from (C) to (D). 



p. 623, interchange lines 26 and 27. 



p. 635, in Table II., to the characters given under (4) add "veinlet dec." 



p. 647, line 19, for "Homoneura" read "Heteroneura." 



p. 650, in last line, the percentage for Hepialidae should be 57.5, and that 

 for Prototheoridae 32.5. 



p. 676, lines 3. 7, and 14, for "lA" read "2A," as in Text-flg. 97. 



p. 682 and ]>. 686, in Text-flgs. 100 and 102, in Iiindwing, for "acl" substi- 

 tute "bc"=basal cell. There is no true areocel in the hindwing. 



p. 685, line 6 from bottom, complete the bracket after "Etischemon." 



p. 689, lines J and 5 from bottom, the percentage for Hepialidae should l)e 

 68.5, that for Prototheoridae 50.0. 

 In Text-tig 110, the forks of Ei + -„ Mii- - and M3 -tt should be hdjelled 



p. 701, line 2. for "M" read "Mr,". 



X. y, z. respectively, to conform with the text. 



p. 707, in Table V (2), under Lepidoptera, for "C" read "D". 



p. 708, in the Phylogenetie Diagram, for "CRETACECUS" read "CRETA- 

 CEOUS", and alter the bracket for "Sialoidea" so as to enclose only the 

 Sialidae and Corydalidae. 



p. 711, line, 13, for "evolutions" read "evolution". 



p. 717, in the explanation of Plate xxxi., fig. 16, for "The arrows point to" 

 read "The arrow indicates the position of". 



I have to thank my friend Mr. A. Philpott, of Invercargill, N.Z., foi- reading 

 through Part 3 \-ery carefully several times, and thus disco\ering a number of the 

 above errors. 



