March 16, 1883.] 



SCIENCE. 



165 



The plan and elevation of this building, which 

 will serve as a type for those installed in most 

 of the lighthouses, is shown in Fig. 7. 



The Planier is a full horizon light. Its 

 characteristic is that of three white flashes 

 separated by a red flash. Its range, like that 

 of all the new lights in the Mediterranean, is 

 twenty-seven nautical miles for fourteen-flf- 

 teenths of the year. 



We have mentioned that the transformation 

 of the Palmyre light is also in progress. This, 

 unlike the Planier, will throw a beam in one 

 direction onlj- ; and the arrangement of the 

 lantern is therefore slighth" diflferent. It is 

 shown in Fig. 8. The general disposition re- 

 sembles, up to a certain point, that of la H6ve. 

 The optical apparatus for the new fixed lights 

 will have a diameter of 0.6 met., instead of 

 0.3, as was formerlj* employed. With the 

 revolving cylinder of vertical lenses, this 

 diameter -will reach 0.7 met. 



CRITICISM OF PROFESSOR HUBRECHTS 

 HYPOTHESIS OF DEVELOPMENT BY 

 PRIMOGENITURE. 



Evolutionists have hitherto been puzzled to find 

 a full and satisfactory explanation of the persistency 

 of certain types, such as tlie familiar Lingula and 

 olliers, through long periods of the earth's past. 

 Prof. A. A. W. Hubrecht of Utrecht has offered, in 

 his inaugural address, an hypothesis which he thinks 

 adequate to solve this problem. The address is pub- 

 lished in full in Nature, nos. 600-691. We may 

 pas.? over the first part, which contains familiar mat- 

 ter only, and which, therefore, we venture to advise 

 scientific readers to skip. The presentation of the au- 

 thor's own views begins near the bottom of the first 

 column on p. 302. The habit of needless diffuseness 

 in writing is a very grave encumbrance to scientific 

 literature, and ought always to encounter the. critic's 

 emphatic condemnation. 



'The theory which Professor Hubrecht has advanced 

 appears to iis not only untenable, but unscientific ; 

 we think it might be characterized as pure specula- 

 tion of that reckless quality which of late years has 

 crept into zoology, considerably to the discredit of 

 the science. To justify this condemnation, we will 

 first state the author's hypothesis, and afterward the 

 objections to it. 



The hypothesis may be summarized as follows : 



1. In many animals the period of reproduction is a 

 prolonged one ; so that there are young born of young 

 parents, others of old parents, and, of course, of par- 

 ents of intermediate age. A distinction therefore 

 exists between first-born and last-born posterity. 



2. Similarly, these first-born will likewise have first- 

 and last-born ; so, also, will the last-born ; conse- 

 quently there will be one set of generations of the 

 first-born, and another set of the last-born. 3. In the 

 first series the generations will follow rapidly, in 

 the second series slowly, upon one another; hence, 

 from a given pair, there will be in time numerous 

 descendants; "'a small number of these being de- 

 scendants in a direct line of the first-born of every 

 successive generation, another small number being 



the descendants in a direct line of the last-born of 

 every successive generation." Consequently, of the 

 contemporaneous generations, the individuals of the 

 first set would have numerous ancestors ; those of 

 the second set, not nearly so many. 4. The age of the 

 parent affects the character of the progeny. Of this, 

 Hubrecht is able to bring forward only one example, 

 — apparently the only one known to him ; namely, 

 that Stone found in the McCloud River that the 

 eggs of young salmon are smaller than those of old 

 salmon. 5. "1 must now call your attention to the 

 second cardinal point. . . . Heredity has, indeed, in- 

 vested them [the progenyl with peculiarities, part of 

 which show themselves in their organization ; another 

 part remaining latent, and only attaining develop- 

 ment in following generations. Such a latent po- 

 tential energy towards eventual modification of the 

 individual or his progeny must needs find more occa- 

 sions to unfold itself in the first-born, simply because 

 these are possessed of a larger number of ancestors" 

 (the italics are ours). 6. Asexual reproduction is 

 accompanied by less variation than sexual. 



From these premises, the deduction: that the first- 

 born of sexual generations are the principal variants, 

 and ergo the principal source of new species; and 

 the last-born, per contra, the representatives of sta- 

 bility. 



In rejoinder to this plausible but specious argu- 

 ment, our contention is, first, that we cannot assume 

 that there are really any series of first- and last-born ; 

 second, that, granting the distinction between them, 

 it cannot be assumed that one is more variable than 

 the other; third, granting both these premises, the 

 facts of zoology cannot be made to show that the 

 permanence of types is derived from the last-born, 

 nor that the evolution of new species depends on 

 primogeniture to any considerable extent. 



First, Any succession of first-born would depend 

 upon both parents being first-born; and the proba- 

 bility of both parents so being for any considerable 

 number of generations is so infinitely small that it 

 might be called zero. Let us take a species which 

 pairs (a bird, for example), and where the male fertil- 

 izes only one female. Let us assume that in a given 

 locality there are ten of each sex, and of various 

 ages, and that there is an equal chance of any two 

 pairing; then the probability of the first-born male 

 pairing with the first-born female would be 1 in 100. 

 The chances of the next set pairing in the same man- 

 ner would be also 1 in 100, if we further assume, 

 what is the usual case, that the number of individu- 

 als remains constant. The chances of both pairs 

 being first-born would be 100 X 100, or 10,000. In 

 nine generations the chance of their being all first- 

 born would become 1 in 1,000,000,000,000,000,000 

 (one million million million). Now, for birds which 

 become mature in one year, these are the chances for 

 nine years. Birds are known first from the Jurassic, 

 which we will call for convenience 1,000,000 years 

 ago ; so that it might prove laborious to write out the 

 chances for that period, the chance being the last 

 term of a geometrical progression of which one mil- 

 lion is the number of terms, and one hundred the 

 ratio. Yet we have taken a case exaggeratedly in 

 favor of Hubrecht's view. It were possible to adduce 

 many arguments to show that the habits of animals 

 often render the existence of a series of first-born' 

 improbable; but the previous calculation sufficiently 

 disposes of Hubrecht's fundamental assumption. 

 And, moreover, every such calculation would lead 

 to essentially the same result, whatever the figures 

 chosen to start with might be, because the chance is 

 the last term of a geometrical progression. If Pro- 



