1 84 



NA TURE 



[December 23, 1897 



emphasising the points on which they chiL-ily dwell. If we 

 would know life we must learn it not from the stage-box or 

 dress circle, but by mingling with mankind. If we would know 

 what scientific work is, we must not be content with listening 

 only to skilfully-told tales of scientific triumphs, but must pene- 

 trate into Ihe observatory or laboratory, where the fear of failure, 

 and the uncertainty of long watching and waiting, are at least as 

 frequent visitors as the a.ssured forecast of success. 



To-day, therefore, I have dealt with problems which are still 

 surrounded by doubt and difficulty, with questions which can 

 only be answered by the combined work of many men, it may 

 be of many generations. It is true that on some of these 

 matters we are gradually acquiring definite knowledge. That 

 earth-air currents, if they exist at all, are very minute in north- 

 western Europe ; that the diurnal variation on quiet days is not 

 necessarily to be regarded as normal ; that local magnetic dis- 

 turbances are due to forces so wide in their range that it is worth 

 while to study them ; these are all facts about which we were 

 in doubt a few years ago, and on which we are in doubt no 

 longer. But greater questions which lie behind these are still 

 unanswered, and if I have ventured to deal with difficulties as 

 well as with certainties, it has been because I have wished to 

 give you a correct idea of current scientific thought on the 

 subject of terrestrial magnetism. 



MR. MERRIFIELD'S EXPERIMENTS 



ON THE RELA TION OF TEMPERA TURE TO 



VARIA TION. 



"POR the last ten or eleven years Mr. Frederic Merrifield, of 

 ■*■ Brighton, has been conducting a most elaborate and 

 extensive series of experiments in the rearing of lepidopterous 

 insects under various conditions of temperature. The re.sults 

 obtained by him are of high interest, both in themselves and 

 also in reference to similar experiments independently under- 

 taken by Dorfmeister, Weismann, Standfuss and others, some 

 of which have already been noticed in the pages of Nature. It 

 is proposed to give here a short general account of the chief of Mr. 

 Merrifield's experiments, with figures of some of the main results 

 obtained. For full details the reader is referred to the original 

 accounts which have appeared from time to lime in the Trans- 

 actions of the Entomological Society of London. 



Experiments in 1887. — The first experiments undertaken had 

 the object of supplying data for an inquiry by Mr. Francis 

 Galton on the subject of heredity. It was anticipated by Mr. 

 Galton that " the experiments would elicit incidentally many 

 interesting results, some perhaps quite disconnected with the 

 objects immediately in view." This anticipation was fully borne 

 out by the facts. 



The first species taken for experiment were two geometer 

 moths, Selenia bilunaria, 'Es^. {il/unaria, Hb. ), and S. tetra- 

 Itmaria, Ilufn. {illustraria, Hb. ), both normally double- 

 brooded in this country. Larvae of both species were reared 

 from eggs laid by females of the spring emergence ; some 

 of the moths resulting from these were selected for pairing 

 according to size, three classes being formed, of maximum, 

 medium and minimum expanse of wing, and the rearing 

 of fresh generations was continued. Some of both species 

 were fed up in the open air ; these showed nothing remark- 

 able. Others (of 6". biiunaria) were kept during all their 

 stages at a temperature of about 80° F., which had the effect of 

 considerably accelerating their development. Five generations 

 of S. bilunaria (counting the moths of the spring emergence as 

 the first generation) were thus produced in the course of the 

 year. These bred moths were all of the summer ox juliaria form, 

 and the females were always larger than the males, which is in 

 accordance with the rule for the natural summer brood. The 

 pairs selected for maximum and minimum expanse of wing pro- 

 duced no fertile eggs after the third generation. The fourth 

 generation consisted entirely of the offspring of one of the 

 medium-sized pairs of the third, and from these a selection was 

 again made as before. The resulting moths of the fifth genera- 

 tion emerged in December and January, showing signs of 

 deterioration. Only one of this brood laid fertile eggs, and 

 these failed to hatch. The average size of the moths increased 

 continuously up to the fourth, but diminished in the fifth 

 generation. 



Experiments in 1888. — The summer of 1888 was cold and 

 wet, and the moths of both the selected species reared from 



NO. T469 VOL. 57] 



larvse kept in the open air showed signs of degeneration. That 

 this was not due to the domestication of their progenitors 

 appeared from the fact that a wild strain of S. tetralunaria 

 behaved in the same manner. 



It was observed in the case of S. bilunaria that the specimens 

 produced from larvce and pupa: that had been kept at about 

 80" F. showed a warmer colouring and fewer spots than those 

 reared throughout their stages at ordinary temperatures. The 

 same was found to be true in a still more marked degree of 

 another species of geometer moih, Ennomos auitimnaria,Vi[&xnb., 

 and also, though to a less extent, of S. tetralunaria. 



Experiments on the pupse of 5". tetralunaria led to interesting 

 results. It was found that moths derived from pupas of the sum- 

 mer brood, first retarded in development by freezing, though ulti- 

 mately forced for a few days, tended, especially in the females, 

 to assume the appearance of the comparatively dark spring 

 emergence. Some larvae of the same summer brood and their 

 resulting pupae were forced from the beginning, with the result 

 that the same batch split into two divisions, the first of these 

 feeding up rapidly and emerging, during the same season, with 

 the summer colouring ; the second taking much longer, and 

 attempting to lie over for the winter as pupae. A continuance 

 of the forcing process brought out some of the latter in 

 November and December, with the summer colouring. Others 

 were exposed out of doors from November 7 to January i, when 

 they were brought into the house and again forced. The moths 

 from this latter group came out in January and February, and 

 were all of the spring colouring. Inasmuch as the continuously 

 forced batch gave the summer form as late as December, it 

 would appear that retardation alone is not in all cases sufficient 

 to determine the assumption of the spring coloration. On the 

 other hand, some pupae of S. teirahmaria reared by Mr. Jenner,, 

 which were similarly trying to lie over for the winter, produced 

 under forcing a series of intermediate forms becoming on the 

 whole more and more like the spring type with the length of 

 time that elapsed before their emergence. Here there was no 

 exposure to winter cold ; but only retardation from constitutional 

 causes. 



Mr. Merrifield remarks that the case oi S. tetralunaria ahovi^ 

 that the alternate succession of the two forms is not a necessary 

 accompaniment of seasonal dimorphism. The same appears 

 from Weismann's earlier experiments, in which, under appro- 

 priate conditions of temperature, the summer form Vanessa 

 prorsa, L., was found to give rise in the next generation to- 

 V. prorsa instead of to V. levana, L. (the spring form). 



The different reaction of members of the same brood of 

 S. tetralunaria to the same conditions of temperature is inter- 

 esting as an example of what may perhaps be called "physio- 

 logical dimorphism " — a principle which there is reason to think 

 is widely prevalent in nature, and which probably favours the 

 survival of those species that exhibit it. 



Experiments in 1889. — These had two main objects, the first 

 being to determine the amount of exposure to cold that could 

 be borne in the different stages, the second to ascertain more 

 definitely the effect upon the perfect insect of temperature con- 

 ditions applied during the immature periods. 



Under the first head it was found that the eggs of both 

 species of Selenia were injured by protracted icing ; a tempera- 

 ture of 80° to 90° F., on the other hand, did no harm and 

 quickened their development. Some eggs of S. tetralunaria 

 gave another good example of physiological dimorphism. 

 Thirty eggs were iced for seventeen days. On their removal 

 from the ice-box, two hatched at once ; none of the rest yielded 

 larvae until from eleven to thirteen days afterwards, when 

 nineteen of them also hatched. 



Icing the larvae of .S". tetralunaria was found to be rapidly 

 fatal, and cooling injurious. The older larvae stood cold better 

 than the young ones. Larvae of S. bilunaria, S. tetralunaria, 

 E. aulumnaria and E. alniaria, L., all endured a continuous 

 temperature of 80° F. or a little more without apparent injury; 

 but one of 90° to 100° F. was very detrimental. 



Further experiments showed that no harm resulted from 

 icing for moderate periods the pupating larvae and pupae of 

 S. teirahmaria and the pupae of E. alniaria, nor from cooling 

 to about 47° F. the pupating larvae or pupae of E. autiminaria. 



Under the second head some interesting conclusions were 

 arrived at with E. autumnaria, E. alniaria and S. teirahmaria. 

 Eggs from a single pair of the first-named species were divided 

 into batches, and larvae and pupae of each batch were brought 

 up under carefully regulated conditions of temperature. The 



