球形特性观察薄片可以通过几种方式形式。1。使不透明——这是玻璃的再结晶。我认为这是更常见的在流纹岩岩石比在玄武岩的岩石。玻璃是一种无机窝无定形的阶段,有一个固态的玻璃过渡到石英和长石组成。球粒的中心通常是某种扰动的玻璃。可能是晶体核或一个小泡沫。这是一个宏伟的球粒流纹岩,从美丽的(集合)[1]Bernardo凯撒:![球粒][2]2。杏仁孔的填充——这通常发生在更基本的岩石,例如玄武岩。一个圆形腔左凝固期间由于气泡的存在。 In this case, the minerals grow from the walls of the cavity inwards, filling the empty space (in contrast to the inside to outside growth pattern of spherulites). The minerals that grow are usually zeolites, clays, quartz and various other hydrothermal minerals. I remember seeing a magnificent example of epidote filling an amygdule once in a basalt. Here's an example of an amygdule from the [Imperial College Rock Library][3]: ![enter image description here][4] Notice how it doesn't completely fill in the gas cavity. These are the two most common ways in which this can form. **However**, in your case I think the answer is less exciting than this. I am pretty sure that your rounded objects are **simply bubbles in the epoxy**. Here's why: 1. They are perfectly rounded. This rarely happens naturally. 2. They are completely empty. 3. The actual rock is to the left of that image. Most of the area covered in your image does not seem to have any real rock in it, unless it is glassy but I doubt it because there is rock on the left. 4. That 4-sided halo around it occurs almost entirely in artificial bubbles created during the sample preparation process. You can put it a λ plate and see that opposite sides have similar colours (either red or blue). [1]: http://microckscopica.altervista.org/en/ [2]: http://i.stack.imgur.com/O5EBP.jpg [3]: https://wwwf.imperial.ac.uk/earthscienceandengineering/rocklibrary/index.php [4]: http://i.stack.imgur.com/xWW6u.jpg