iiMrkiiuis and attUades of h'pklopfcrons hirvce. 301 



&c. There is one fact about the formation of the 

 obhque stripes that strongly confirms this opinion. It 

 has ah-eady been pointed out that the 7th stripe is 

 always the first to have its ground colour altered to the 

 tint of the dots, and the first part of the stripe to 

 undergo this change is that which crosses the 8th 

 abdominal segment and enters the base of the horn. 

 Now this is the only segment crossed by the oblique 

 stripes which is not annulated. All segments anterior 

 to the 8th abdominal are divided into rings (generally 

 eight in number), and upon these shagreen dots are 

 arranged in regular rows, one upon each ring. It 

 follows that the rows of dots are separated by the deep 

 furrows that intervene between the rings, and the 

 enlarged dots that form the oblique stripes are therefore 

 similarly separated. Hence these dots are rendered 

 distinct, and the constitution of the stripe is quite 

 apparent, when fusion would have taken place (pro- 

 ducing quite different appearances), if the segments 

 were not annulated. But the 8th abdominal segment is 

 not annulated, and here therefore the spreading area 

 round the base of each dot has not to cross a furrow in 

 order to coalesce with that of the next dot. Consequently 

 fusion takes place at an early date, and this stripe is 

 generally much larger and more distinct than the 

 others, being the oldest continuous stripe in the indi- 

 vidual ontogenj^, and doubtless in the phylogeny also. 

 Some appearances led me to suspect that the stretching 

 of the larval skin during growth in each stage was one 

 factor in causing the dots to spread and fuse. It was at 

 all times obvious that the dots were most entirely 

 separate immediatel}' after ecdysis. Again, the rapid 

 disappearance of the small dots upon the back, at the 

 beginning of the fifth stage of *S'. lir/ustri, seems to be 

 chiefly due to this process. 



The importance of the shagreen dots in this respect 

 is seen in the fact that most of the larval markings of 

 all the English species of SvierintJivs, of Maerofilossa, 

 and of SpJtiiix ligusiii, are due to their arrangement, 

 size, and fusion (in many cases). The same thing is 

 probably true of many other Sphimiidie, but they have 

 not hitherto been examined from this point of view. 



Since writing the above I find thatWeismann attributes 

 importance to the shagreen dots in relation to markings 

 in Ddlephilu euphoihke and V. Idppopha'es, suggesting 



