A long standing discussion on the existence of large raindrops has received further impetus from observation of occasional large drops 4-8 mm. diameter in Hawaii (Rauber et al 1991). Such drops can only exist in the absence of collision with smaller drops which lead to break up, so their existence only occurs in the absence of significant numbers of intermediate size drops, and without any equilibrium spectrum from coalescence and break up (McFarquhar and List, 1991a, b). It is hypothesized that large nuclei originate a few drops which are recycled to produce the low concentration of large drops. The ice phase is evidently absent, so no break-up inhibition by ice occurs, as could be the case in cold clouds.
Laboratory studies, without electrical effects, of smaller drop axial ratios and fall trajectory has confirmed the role of eddy shielding in drop distortion by individual oscillation for droplets near equivalent diameter 1.1 mm. A maximum in the sideways drift occurs near this size, (equivalent to 6% of the terminal fall speed) having lower value (2%) for both smaller drops (1 mm,) and larger drops (1.54 mm), (Beard et al 1991, Beard and Kubesh 1991). The variance in axial ratios is greatest for the maximum size examined (1.6 mm), and demonstrates a broader size coupling between eddy shedding and drop oscillation than for smaller sizes. Such effects could extend to larger sizes. An elegant distortion technique is used to examine the oscillation mode. The time spent by oscillating drops in vertical extension influences radar polarization back scatter. Larger drops (2--2.5 mm diameter) were similarly examined after a seven storey fall (Kubesh and Beard 1993). The larger drops showed a larger mean dispersion around equilibrium shape, resulting from transverse fundamental and first harmonic oscillation caused by side eddy shedding.