May II, 1888.] 



SCIENCE. 



229- 



lack at present, and at the same time to develop the resources of 

 the country. In order to make the work of such a survey as useful 

 as possible, it is proposed to make the results known through the 

 agency of the daily press and other publications, to be issued as 

 rapidly as possible. It is to be hoped that the practical and wise 

 measures proposed by the association will be carried out, as they 

 cannot fail to benefit the people of the State. 



LETTERS TO THE EDITOR. 



•,' Corres^oiiients are requested to be as brief as possible. Tlie writer's name is 

 in all cases required as proof of good faith. 



Tw.'tty copies of tin nu mber containing liis communication ivill be furnished 

 free to any correspondent on request. 



T/ie editor will be glad to publish any queries consonant with the character of 

 the Journal. 



Is the Rainfall increasing on the Plains ? 

 In view of the recent discussion on this subject which has ap- 

 peared in this periodical, perhaps it will not be amiss to add a few 

 remarks to those of Mr. Curtis concerning the errors which may 

 arise in climatic studies from errors in rainfall records. Rainfall 

 records have probably been kept as long in New England as in any 

 other part of the United States, and a number of them give indica- 

 tions of a secular change in the amount of rainfall. But Mr. E. B. 

 Weston, Desmond Fitzgerald, and others who have had occasion 

 to study some of these records, state that in certain cases the ap- 

 parent change in the amount of rainfall was clearly due to the dif- 

 ferent methods pursued by different observers in measuring snow, 

 and converting it to its equivalent in rain ; and they think but little 

 reliance can be put in the results obtained from a comparison of 

 earlier with more recent records. Mr. Weston has also shown that 

 gauges with different kinds of rims give persistently different re- 

 sults. If these early volunteer records are of uncertain value for 

 studying climatic changes, are those which are now being gathered 

 by our national Signal Service to prove more so ? These latter 

 records are in almost every case obtained from rain-gauges e.xposed 

 on the roofs of houses; and hence the amount of rain caught 

 becomes a function of the wind-velocity, a function of the wind- 

 direction, and a function of other variants and variables, not least 

 among which is a not uncommon change in the position of the gauge 

 itself. Numerous experimental observations have shown that gauges 

 exposed on roofs catch more rain when exposed on the side opposite 

 to the direction from which the wind blows, and less rain when ex- 

 posed on the same side from which the wind blows. The writer 

 thought that the large errors which may arise from this source were 

 fully recognized by the Signal Service officers and by the scientific 

 public, so that it was unnecessary to call attention to them. But re- 

 cently he has listened to two papers by well-known writers, dealing 

 with changes in the amount of rain, especially in the West ; and both 

 these writers referred to the valuable records now being obtained 

 by the Signal Service as furnishing a basis for future studies of this 

 kind. The present writer inferred from these papers that the errors 

 arising from exposure are not so fully known as they ought to be, 

 and hence presents a brief study of the Boston rainfall record, which 

 is only one of several similar cases which have come under his 

 notice. For several years in succession the annual rainfall at the 

 Boston Signal Service station has been reported below the normal 

 According to the Bulletin of the New England Meteorological So- 

 ciety, in 1885 it was nearly three inches below the normal, in 1886 

 nearly five inches, and in 1887 nearly thirteen inches below. This 

 seemed rather strange, since none of the numerous gauges around 

 Boston showed such marked deficiencies. Thus, in 1887, when the 

 Boston Signal Service station reported the annual rainfall thirteen 

 inches below normal, the Harvard College Observatory, only three 

 miles west of Boston, reported an annual rainfall twelve inches 

 greater than that reported from Boston, and one inch greater than 

 the average of twenty years' observations at the observatory. The 

 observer at Lynn, Mass., ten miles north of Boston, reported an 

 annual rainfall fifteen inches greater than Boston, and six inches 

 above the average of thirteen years' observations at Lynn. Accord- 

 ing to the records of several gauges in Milton, ten miles south of 

 Boston, the annual rainfall was from nme to twelve inches greater 

 than at the Boston station. These stations are all so close to 



Boston, that it is rendered entirely improbable that there was in. 

 reality any great deficiency in the Boston rainfall ; and the apparent 

 deficiency seems clearly due to a change in the position of the Bos- 

 ton gauge about 18S3 or 1S84. Previous to this the gauge had 

 been exposed on the roof of the Equitable Building in Boston, and 

 these records were used in forming a series of averages or normals.. 

 Then the gauge was removed to a high tower on the Post-Office Build- 

 ing, and since then there has been almost a persistent deficiency of 

 precipitation as compared with former records, or with the records of 

 stations surrounding Boston. Moreover, the amount of rainfall caught 

 is evidently a function of the wind-velocity, and decreases with in- 

 creased velocity of the wind. Thus, during a gale on April 3, 1887, 

 the amount of precipitation reported from the Boston Signal Service 

 station was 0.22 of an inch ; while measurements by a number of 

 observers in and around Boston showed that snow fell to a depth 

 of over a foot, and when melted gave an inch of precipitation as or- 

 dinarily recorded. -Again, during the storm of March 11 to 14, 

 1888, the Boston Signal Office reported 1.24 inches of precipitation,^ 

 while surrounding stations reported three inches or more. 



It seems a pity that our Signal Service gauges should be so badly 

 exposed, for these are looked to as the standards throughout the- 

 country ; and there is no doubt that in the future, as in the past,, 

 there will be attempts to proveclimatic changes from their records ;; 

 but the writer feels that anyone who has had experience with rain- 

 fall observations will look dubiously on any conclusions based on, 

 such records as exist at present. H. HELM Clayton. 



Blue Hill Observatory, May 1. 



Significance of Sex. 



Some recent publications on the subject of the significance of 

 sexual reproduction, especially those of Dr. Weissmann {Nature, 

 xxxiv. p. 629, 1886, and xxxvi. p. 607, 1887) and a short abstract of 

 a lecture by Hatschek {A7inah and Magazine of Natitral HistOiy, 

 i. p. 163. 1888), have induced me to draw brief attention to some 

 speculations of my own on this subject, published several years, 

 ago. 



I. Dr. Weissmann, in his admirable paper on the significance of 

 polar globules {Nature, xxxvi. p. 607, 18S7), after showing that 

 there must be some very great benefits resulting from the introduc- 

 tion of sexual reproduction, says, " Such beneficial results will be 

 found in the fact that sexual propagation may be regarded as the 

 source of individual variability, furmshmg material for natural 

 selection." Now, in an article on genesis of sex, published in the 

 Popular Science Monthly, December, 1879 (xvi. p. 167), and repub- 

 lished in the Revue Scicntifiquc for Feb. 14, 1880 (xviii. p. 220), the 

 same thought is distinctly implied, though not distinctly expressed. 

 The whole contention of the article is to show that the object of sex 

 is the funding of individual differences in a common offspring,, 

 thereby improving the offspring ; and, further, to show how much 

 pains nature has taken to make individual sexual differences greater 

 and greater in the history of evolution. In the last paragraph I 

 say, " Such mixing produces mox& plastic nature, more generalized 

 and therefore more progressive form." 



This was written nearly nine years ago. Meanwhile the thought 

 continued to develop in my mind. In a book (' Evolution and its 

 Relation to Religious Thought') just now published, but most of 

 which, and especially all on this subject, was written three years 

 ago, the same thought is much more distinctly expressed. On p. 

 220 I say, " Why was sex introduced at all } There are doubtless 

 sufficient reasons of many kinds, but X\\e fundamental reason con- 

 7iected with evolution is the funding of individual differences in a 

 common offspring, thereby giving to the offspring a tendency to 

 dive?gettt variation." Again on p. 223 : " Complexity of inheritance, 

 like complexity of composition in chemical substances, gives insta- 

 bility to the embryo and liability to variation to the offspring; and 

 this in its turn furnishes material for selection of the fittest." This, 

 was written in the fall of 1884 ; but, being much pressed with other 

 work at that time, I laid aside the manuscript, and only took it up- 

 again, finished it, and sent it to the publisher, about a year ago. I do 

 not bring this forward now by way of reclamation, — for even if I had 

 any right to make such, which I have not, I care little who brings 

 out a truth, — but partly because I would not seem to borrow aa 



