No. 2, November, 1920] GENETICS 95 



bircli and alder and inbred for 5 seasons, filigrammaria showed no characters 'if the birch- 



alder-Iarch-eat ing autumnata. 



A local larch-pine race of autumnata has developed in Wilton Wood, Yorkshire, since 

 moorland reforestation about 1X1)0; a neighboring isolated larch forest (Normanby Intake) 

 was destroyed in lNSf> and birches came in. Larch-feeding insect of Wilton is smaller, duller, 

 feebly marked, a month earlier than the more typical birch-feeding Normanby form. But 

 the birch inst inct lias not been lost in t lie Wilt on variety. Birch feeding restores original size, 

 though natural selection by bats, owls and night-jars has tended to eliminate pah;, silvery 

 variants in favor of darker and feebly-marked. Larlier emergence of larchwood race is due 

 to gradual fall of temperature under extremely cool moist conditions of the dense larch forest 

 in contrast with dryer, more open birch woods. In evidence, pupae exposed to a gradual fall 

 in low temperatures emerged earlier than others exposed to a similar fall at higher tempera- 

 tures. Caterpillars of larchwood stock fed on birch produced imagines indistinguishable 

 from their own race in color and early time of emergence (1), but of larger size. 



A local brilliant silvery variety of O. dilutata is found in an isolated oak wood cut off by 

 a ridge 1000 feet high from the nearest oaks, f mile distant, where "the ordinary suffused 

 melanic form" occurs. — The increasing prevalence of melanism of moths near English indus- 

 trial centers is ascribed to "changed metabolism favoring resistance to, or actually induced 

 by, the use of food contaminated with metallic salts and other compounds. Exhaustive day- 

 light observations on Polia chi showed that no natural selection favorable either to dark or 

 light forms occurred, although natural selection is not entirely ruled out. — Melanism in Opo- 

 rabia is non-Mendelian. A persistent blend occurs. — A 9 mutation of autumnata, "lati- 

 fasciata," with barred wings, behaves as a sex-linked (partial) dominant, heterozygous for 

 sex. Crossed with recessive type d\ a sex criss-cross results (cf c? heterozygous, latifasciata; 

 9 9 recessive type). Heterozygous dominant Fi d 1 (Z'Z, using Morgan's symbols, Z' being 

 chromosome carrying latifasciata gene) mated with the recessive type 9 (WZ), either of Fi 

 or of wild stock, gave heterozygotes and recessives of each sex in equal numbers (latifasciata 

 9 9 and dV + type 9 9 and dV, i.e., WZ' + Z'Z + WZ + ZZ). Recessive F x type 9 mated 

 with type cf bred true. — Crosses between O. autumnata and dilutata are probably sex-linked, 

 though it is suggested that, since the latter has only 30 pairs of chromosomes instead of 38, 

 "mitotic anomalies" may explain the phenomena. Great disturbances in development 

 occurred: (1) 9 9 from autumnata 9 X dilutata d 1 emerged three months before the males 

 and had no ovaries, (2) dilutata 9 X autumnata d 1 produced few females, never more than 

 one per brood, but fertile and like autumnata. Lack of viability is manifest in pupa?. Though 

 both reciprocal Fi hybrid cfcf are fertile, back-crossing never produced adults. No synapsis, 

 nor reduction division, occurs. A larva with wings appeared in Fi ex autumnata 9 X dilu- 

 tata d 1 . Bacterial disease delays gametogenesis and hence emergence, but precocious ovari- 

 less 9 9 of 0. autumnata 9 X dilutata a" , though infected, appeared at their accustomed time. 

 — Hybrids between autumnata and filigrammaria are a blend, no segregation appearing in F2 

 nor in back-crosses, but in F 3 "a delayed or pseudo-segregation is manifested in which a 

 portion of the brood still remains intermediate whilst the remainder appeared in a uniform 

 pseudo-mutational guise," recalling behavior of Oenothera Lamarckiana. No segregation of 

 marked size-differences of parents in support of multiple factor theory. A female with three 

 antennae appeared in intermediate portion of F 3 lot. Back-crossing Fi (blend) with autum- 

 nata produced in first generation autumnata, but in back-crossing with filigrammaria two gen- 

 erations are necessary to bring hybrids back to pure filigrammaria facies. Reduction division 

 of Fi filigrammaria 9 X autumnata d 1 , involving 37 X 3S chromosomes (haploid numbers) 

 is nearly perfect. Crosses between filigrammaria and dilutata failed, apparently through 

 lack of physiological affinity. Ova from Cheimatobia boreata 9XO. autumnata c? failed 

 to hatch. No ova from reciprocal cross, nor from C. brumata X O. autumnata. — John H. 

 Gerould. 



678. Heal, John. Streptocarpus hybrids. Gard. Chron. 67: 293. June 12, 1920 — 

 Breeding of Streptocarpus is modern, but many advances have been made in late years. Green- 

 house races originated from Kew hybrids. First flowers were comparatively small, on short 



