July 1S91.J 



PSYCHE. 



113 



a word must be said about the technique em- 

 ployed by Prof. Graber. As he informs us, 

 and as is quite evident from his twelve large 

 colored plates, his results were largely ob- 

 tained from germ bands isolated from the 

 yolk, stained in toto with carmine, and 

 mounted in some resinous medium. His 

 first plate shows that he has also examined 

 Lina embryos unstained and in situ on the 

 yolk. Now both of these methods, though 

 useful for some purposes, are quite inade- 

 quate to decide any delicate question con- 

 cerning surface relief, and have consequently 

 been all but abandoned by some recentworkers 

 In point of detail Prof. Graber's surface 

 views of Lina cannot stand comparison with 

 some of the figures of insect embryos pub- 

 lished decades ago, while nowhere is the in- 

 adequacy of his isolation method better 

 shown than when he attempts to elucidate 

 the structure of the brain. When repre- 

 sented at all in his figures this important or- 

 gan is incorrectly represented. In order, 

 therefore, properly to appreciate Prof. Gra- 

 ber's observations it is necessary to bear in 

 mind that his technique is somewhat defective. 



In the 15 quarto pages devoted to the ner- 

 vous system there are many new and inter- 

 esting facts, but we miss a careful treatment 

 of the very earliest stages in the formation 

 of the ganglia, both cephalic and ventral. 

 It is safe to say that a few good sections 

 through the nerve-cord of a sufficiently young 

 Stenobothrus embryo would have brought 

 out some interesting facts on the formation 

 of the median and lateral cords — facts which 

 would have induced the author to view the 

 nervous system of the Coleoptera in a little 

 different light. 



A short time ago Prof. Graber devoted a 

 paper to the important subject of metameric 

 segmentation in insects. According to the 

 observations therein recorded, the first seg- 

 ments to make their appearance in the em- 

 bryo are not the definitive body-segments 

 {microsomites), but segments nearly or quite 

 corresponding to the imaginal aggregates of 



segments (head, mouth-parts, thorax, abdo- 

 men) ; the definitive segments being formed 

 by a splitting up of these macrosomites. Al- 

 though it occurred to Prof. Graber at the time 

 that this phenomenon might be due to a fore- 

 shadowing of adult structure, he chose to 

 adopt the view that the early macrosomitic 

 segmentation was an ancestral feature. In 

 his present paper he devotes considerable 

 space to this subject, bringing out quite an 

 array of pseudo-mathematical formulae, and, 

 notwithstanding Heider's very sensible com- 

 ments on his former paper, still persists in 

 seeing some mysterious palingenetic trait in 

 macrosomitic segmentation instead of an 

 anticipation of the ultimate adult structure. 

 That the latter is the correct explanation is 

 shown by a study of Xiphidium. In this 

 Locustid the definitive segments make their 

 appearance in a wave which runs from the 

 anterior to the posterior end of the germ- 

 band. When the whole postoral portion of 

 the germ-band has thus been split up into 

 about eight segments, the remainder of the 

 definitive segments are successively intercal- 

 ated just in front of the caudal plate. Then, 

 and not till then, does macrosomitic segmen- 

 tation set in. Although this method of 

 growth by intercalation of segments in front 

 of the anal plate has been repeatedly shown 

 to be the typical method in Annelids, Crus- 

 tacea, Peripatus, Arachnids, Myriopods and, 

 to a certain extent, in Hydrophilus (Heider), 

 Prof. Graber maintains that it does not occur 

 in insects. Strangely enough the very fig- 

 ures of Stenobothrus to which he appeals, 

 prove the very opposite of his contention, for 

 they show quite clearly that the youngest 

 segment must lie just in front of the anal 

 plate. The first indications of segmentation 

 have probably escaped Prof. Graber, — it 

 being impossible, as we have found after 

 repeated trials, to detect them in semi-trans- 

 parent, isolated germ-bands. 



Prof. Graber divides insect embryos into 

 microblastic and macroblastic, or long and 

 short ones. Stenobothrus is microblastic; 



