12 



KNOW I.I. nr.i:. 



Januauv, 101.?. 



Moiv rarely, such a inctlini^ causes the secomlary 

 rays to uniliTKo (Irllcctioii. aii<i Id f^row thereafter 

 in entirely lu-w ilin'ctioiis. as so intt'rcslin(;l\' sliown 

 in r'i(,'iire H. 



In rare ca.ses, siicii a nieetiii},' ina\' even cau.se 

 curvilinear deflect ion. 



lCs|H'cially mysterious and interesting are those 

 cases in which secondary rays fail 

 to form along the primary' rays, s. ■ 

 Ixautifully shown in I'igure 10. 



The mystery is heightencil. because 

 in njost cases, and even in this. :i 

 close inspection of the edges of tli< 

 primary rays shows that priiniti\c 

 Secondary rays formed along the 

 primary ones, hut were suhsefpieiitly 

 either encased in soliil layer growth, 

 or failed to increase appreciahK . 

 .\nother mystery concerns tin 

 occasional haliit some of the second- 

 ar\' ravs have, of growing downwaril 



or inward, instead of in the usual 1 loi kl lb 



outward manner. This droo[)ing 

 habit is exquisitely shown in Figure 1.5. 



More puzzling still arc those crystals in which, for 

 no aj^parcnt reason, secondary rays form only in 

 alternate order, and upon only three of the main 

 ravs, producing a trigonal effect •^o I'vnphirnlK 

 jiictured in Figure 17. Odd freak 

 crystals, due in some cases to 

 accident, in others seemingly to 

 design, are not uncommon among 

 the snows of some general storms. 



Those of a trigonal shape. 

 Figure IS, those exhibiting a ten- 

 dency to divide into four or eight, 

 and the odd three or four vaned 

 crystals, and some others, belong 

 in this category. Those havin.i; 

 four segments are shown in 

 Figure 19. The most strange ant! 



rare freaks are, perhaps, those Kigcrk 26 



that undergo a change of type from 

 trigonal to hexagonal, or hexagonal to trigonal, as 

 growth progresses, as shown in our Figures 20 

 and 21. 



Among the more strange freak tabular crystals are 

 those of chaotic design. Figure 22 is a case in point, 

 in which the several parts are not only unlike and of 

 uneijual sizes, but are also attached in irregular order 

 around a central, and irregular, nucleus. Crystals 

 of this character seem to form most frequenth' 

 within eastern storm segments, when tabular crystals 

 lirst appear among columnar ones, ])reliminary to a 

 complete or partial change of tyiie to the former. 



Possibly the mf)st singular type due to design, 

 is the comi)ound "cuff button tyiie" (Figure 2.5), due 

 to tabular outgrowths occurring from the ends, and, 

 more rarely, sides, of columnar crystals. 



Doubtless in many, jjeriiaps all. such cases, the col- 

 umnar part formed Hrst at high and cold altitudes, 

 and the tabular parts later, at some lower cloud level. 



n 



.\mong the more puzzling forms are tlmsi- 

 exhibiting binate symmetry, and in which one-half 

 or two-thirils of a crystal differs from the other 

 portion, as in I'^igure 24. The crystals possess such 

 an iimazing richness and complexity of interior that 

 scant mentif)n of these features can be made in ;i 

 single article. Perhaps the most wonderful and 

 inexplicable of these interior features 

 are th(; tiny sets of clustered micro- 

 scn|)ic air bidibles which some few of 

 them ])ossess. These usually en- 

 circle the nucleus, and arc arranged, 

 in some cases, in a symmetrical order 

 ])erfect almost beyond belief, as 

 shown in Figure 25. 



Accidental causes produce many 

 freak crystals. Crystallization often 

 occurs w bile the cr\stals are crowded 

 close together, or lie partly embedded 

 in a cluster, so that .some parts have 

 free, while others have scant, oppor- 

 tunity of growing. Crystallization, 

 moreover, often occurs upon fracturcil 

 cr\stals, upon broken parts, all of which cau.ses tend to 

 produce irregular forms, oblong crystals, and so on. 

 The latter are oft-times csijeciallv interesting. as the\- 

 show very beautifully the attempt of the crystallizing 

 tMicis to Imild up symmetrical structures around 

 imperfect nucleii. 



Very interesting are those cases 

 in w hich two or more crystals grew, 

 in part, while lying close together. 

 riie closely lying parts having, of 

 course, less material available and 

 su[)plied them than the outlying 

 ones, necessarily grow in a stunted 

 manner. In many cases, branches 

 or other adornments actpiired else- 

 wiiere, failed to form thereon, 

 I ausing a break in the continuity 

 (if the design. 



In many cases, however, the 

 cr\stals seem to attempt to repro- 

 duce the general pattern clear around, even though 

 sup[)lied with insufficient material at some points, 

 and do so to the degree that the design, though 

 thinner and smaller, is yet kept, in part at least, 

 intact, as shown in Figure 26. The more perfect 

 crystals of snow are doubtless the most perfect and 

 beautiful examples of Nature's inorganic art and 

 of symmetrical design occurring in Nature. Their 

 value, whether in an educational way as objects 

 for nature and crystallographic study, or in the 

 realm of art, as models for designs in jewellery, metal, 

 wall [laper, lace, china, silk goods, and so on. should 

 be very great, and has led already to their being 

 extensively used by universities, museums, schools, 

 lecturers, art craft shops, designers, and so on. 

 There will doubtless be a great increase, in the 

 future, in the uses to which they will be put, as 

 these bits of pure beauty from the skies become 

 more gener.dly known and ninrc fully a[)preciated. 



