Three species and one subspecies of fleas, all belonging to the Ceratophyllidae, are described and figured in this paper. One of the species, formerly known only from Siberia, and imperfectly described, is redescribed, and recorded from Yukon Territory. The remaining two species and the subspecies are believed to be new to science.

Reproduction in the Pseudococcidae is usually regarded as being predominantly, if not exclusively, bisexual, though Ferris (1950) has pointed out that within the family as defined by him males are known for only a small percentage of the species described to date. Detailed biological observations have been restricted to those members of considerable economic importance and in regard to reproductive habits there are few specific references in the literature. Hough (1922) in his account of the life history of the clover root mealybug Chnaurococcus trifolli (Forbes) Ferris, stated that no males were seen throughout the period of the observations. James (1937) made a comprehensive study of the reproduction of Planococcus citri (Risso) Ferris, Pseudococcus maritimus (Ehrh) Ferris, P. gahani Green, P. adonidum (Linn.) Ferris, and Trionymus peregrinus Green, and found that parthenogenesis did not occur in any of these species. This confirmed conclusions of earlier workers with Pl. citri (Schrader 1923, Sutherland 1932), and with P. gahani and Phenacoccus gossypii Twns & Ckll. (Basinger 1934). However, Myers (1932) reported parthenogenesis in some of his experiments with Pl. citri but his procedure of isolating females from the cultures after the second moult is clearly open to criticism.

The following is a description of a new species of leafhopper from Eastern Canada of the subgenus Cloanthanus of the genus Scaphytopius. A key to the known Canadian species of Cloanthanus is included. Two of these species, angustatus (Osb.) and argutus DeL., do not appear to have been recorded previously for Canada. The key is based primarily on structures of the male genitalia, as these show the most reliable specific characters. External characters of colour, markings, and structure are often unreliable, because of the variation that occurs. There is occasionally some variation in the genitalic characters, notably in the degrees of curvature of the paraphyses, but the general shape of each structure is sufficiently constant to provide reliable specific characters. This subgenus, which was revised by Hepner (1947), is poorly represented in Canada, as the majority of the species have southerly distributions in North America. The common Canadian species is acutus (Say), followed by latus (Baker); the remaining species are more restricted in their ranges and distributions in Canada.

The seed production of many field crops, and in particular of legumes, depends to a large degree upon the number of pollinating insects present in the field during the blooming period. A study of the various factors affecting seed production requires, among other things, a knowledge of the population of pollinating insects present, and a ready means of comparing pollinator populations from one field to another. To obtain this information some means must be employed to count and classify the pollinating insects in the field.

General.—The purpose of this paper is to analyse a sampling method devised to assess larval populations in an outbreak of the lodgepole needle miner, Recurvaria milleri Busck (Busck 1914, Hopping 1945).

The problem of developing an adequate sampling method is intimately concerned with the life-history of the insect, the region of the outbreak and the nature of the forest stand in which the outbreak occurs. In sampling most defoliator populations the problem is made more difficult by external feeding and wandering habits, hence it is usually done in some relatively inactive stage at a fixed time. de Gryse (1934) describes the problems inherent in sampling these insects. The needle miner, however, is fixed in its location for most of its life-cycle and is therefore readily obtainable for study. The problem here is reduced to a statistical one, that of obtaining an acceptable sample i.e. within suitable error limits with due regard for existing variables.