Claudius Ptolemy · 克劳狄乌斯·托勒密
Almagest · 至大论
Ptolemy and the Anomalies of Venus
Paper / 论文 · Date from document metadata / 日期据原稿文件信息 · Editorial title / 本版编拟标题
To fulfill what he lays out as the mathematician’s task, Ptolemy will have to explain away the apparent anomalies incurred by Venus via uniform circular motions. The two anomalies are characterized thus in Chapter 2 of Book IX: “one of the anomalies is seen to have relation to the sun, and the other to the parts of the zodiac”. Later, when the motion of Venus is analyzed, this preliminary distinction is delineated in a clearer and more careful way. What the eye observes are, first, Venus sometimes appears to the sun’s west, sometimes its east, seemingly following a periodic pattern, and second, the largest possible angular distances (i.e. largest elongation) between Venus and the mean sun (the difference between the apparent sun and the mean sun is negligible) varies in relation to the location where the complex (Venus in relation to the sun) appears.
In other words, the first anomaly has to do with the fact that as Venus circles around Earth with an average speed equal to the sun, it sometimes overtakes the sun, while at other times falls behind. The second anomaly, on the other hand, states that at two different positions along the circuit traveled by the complex, the largest elongations ever possible of Venus from the sun at those positions will also be different. Here, what I mean by the largest “possible” elongation is that, as the complex is circling around Earth throughout time, it shall reach the same position at every cycle, and Venus in turn shall change its position on its epicycle with respect to the sun every time the complex reaches that position, and Venus will eventually reach the position on its epicycle so that it appears to the eye to be farthest from the sun, and this position constitutes the largest possible elongation. Yet for this second anomaly to stand, it must first be assumed that there is a largest possible elongation of Venus from the sun at any position. This is a hidden assumption that is not carefully established by Ptolemy, for if we were to make the speed that Venus travels on its epicycle the same as the speed that the epicycle itself travels on the deferent, we shall see that except for two positions, there is no position in which Venus will be at its largest possible elongation from the sun, but that when the complex reaches a certain point on the deferent, Venus should also be at a determined position on its epicycle from the sun. Or if the speed of the complex traveling on the deferent shall be a multiple to that of Venus on the epicycle, it is also impossible for there to be a largest possible elongation for any given position. Thus, there is this huge assumption which Ptolemy does not clarify that allows for the later presentation.
Once these two anomalies are clarified, it seems possible to account for them right away, using the extant hypotheses Ptolemy uses to explain the movement of the sun. However, Ptolemy states in Chapter 5 of Book IX, that “immediately on applying the particular positions observed to the courses constructed from the combination of both hypotheses and continually examining them together, we find things cannot proceed so simply”. The major problem he sees in these observations is that, even if a combination of the epicyclical and the eccentric hypotheses is used, there seems to be one other anomaly remaining to be solved.
How does a combination of the two hypotheses solve the first two anomalies? Venus being sometimes to the sun’s west and sometimes to the sun’s east can be accounted for by Venus’ own movement on the epicycle. Since the center of the epicycle moves with a speed equal to that of the mean sun, and the mean sun is not far from the apparent sun, one could say that, when Venus appears to be on the west side to the sun, it is because it has moved to one side of the epicycle, and when it appears to be on the east, it is in turn because it has moved to the other side of the epicycle. The second anomaly can be explained by the eccentric hypothesis, for according to the second hypothesis, Earth is placed slightly off center in the circle on which the complex moves. The effect of this is that one side of the deferent will be farther away from Earth than the other side. Then reasonably on one side of the deferent the complex will appear to the observer larger than when it is on the other side. Therefore, on one side, the largest possible elongation of Venus to the sun will be larger than the largest possible elongation on the other side. By this combination of the two hypotheses, the first two anomalies are resolved.
The third anomaly, it seems to me, is only generated after the introduction of this combined model. Ptolemy talks about it as a discovery from trial and error, but refrains from pinning down its exact nature. Inferring from the consequences the resolution, i.e. the introduction of the equant, brings about into this combined model, one could say that the third anomaly is the complex moves on the deferent by a non-uniform rate. Yet this anomaly would not be fully articulated if not for the pre-establishment of this combined model, by which the complex itself is constructed.
Now, the first two anomalies plus a generated third anomaly can all be resolved theoretically, thus what remains for an account of Venus’s movement to be completed is to find its starting point in observations. What one needs first, then, is a collection of data that records those various largest possible elongations at different positions.
For we see from our study in geometry, that Venus would be at the largest possible elongation from the sun when it is at the intersection between the tangent line from Earth to the epicycle and the epicycle itself, for it is at tangent that the angular distance shall be the greatest. And we also know that geometrically the closer is the complex to Earth, the greater shall be its largest possible elongation as it appears to the observer. And the closest it gets to Earth is when it is on perigee, the farthest away on apogee. Therefore, we could find the apogee and perigee on the deferent through finding where the largest of the largest possible elongations appear. As the line of apsides is determined, so can all the data be ordered on the model. Moreover, by the data one already obtains from observations, the ratio between the radius of the epicycle and that of the deferent can be determined as well, for by following Ptolemy’s steps in geometrical demonstration, one is capable of determining the respective relationships between the linear distances Earth has to the center of the epicycle, the diameter of the deferent and the radius of the epicycle. By these, one is also capable of determining the eccentricity between the center of the deferent and Earth. What is interesting in this procedure is that when Ptolemy is harmonizing the results he obtains respectively from when the complex is at apogee and when the complex is at perigee, he employs a method which one might tentatively call algebraic. For he first assumes one thing (the linear distances from Earth to the center of the epicycle) to be known, and thus makes the thing sought for (the radius of the epicycle) known; when the results on two sides differ, he harmonizes them by establishing two ratios that share one common term. Yet one has to be careful here and see that his method is exactly not algebraic: instead of assuming the thing sought for as known, he makes assumptions only for the starting-points. In other words, his procedure is standard in the sense that he sticks to the formula of positing hypothesis, and starting from the hypothesis reaching for the result, rather than posit the result as something already known.
That said, Ptolemy moves on to determine the eccentricity between the equant, i.e. the center around which Venus moves uniformly, and Earth. He achieves this by selecting a specific position, as he says in Chapter 3 Book X, “when the sun’s mean position was a quadrant’s distance in one direction or the other from the apogee”. Yet, after the data are gathered from observation, he says, before the demonstration starts, that “for the proposed mean position, the greatest morning elongation is assumed to be ... and the evening one ...” (my emphasis and omission)
By this, he is addressing the problem that the largest morning and evening elongations appear unequal. The apparent inequality comes from this third anomaly that is to be resolved by the equant. What is to be noted is that, his interpretation of the data collected by earlier observers here seems to be that the mean sun assumed by them has Earth as the center of movement, and not the deferent. Hence he conveniently arrives at the conclusion that as the mean sun in his model moves a quadrant away from the perigee, so is the mean sun assumed by his predecessors at direct quadrature to Earth. In other words, against the backdrop of Ptolemy’s own template, the data collected by the earlier astronomers appear anomalistic, which results from their hidden assumption that the mean sun moves uniformly about Earth. Only by this interpretation is Ptolemy able to carry out the further calculations, which add the two elongations up to compose of a larger angle, which could decompose into two equal angles, each of which is the elongation. Or, it might not be his “interpretation”, but he intentionally selects those data collected by those who he knows carry such an assumption, to lighten the weight of his labor. Otherwise, if the mean sun is already supposed to move on an eccentric circle, the direct parallelism between the movement of the assumed mean sun about Earth and the movement of the mean sun that fits his model about the center of its movement will be lost (On the diagram on page 315 in Taliaferro’s edition, the parallelism is between the line DE and the dashed line). Without the direct parallelism, the angle between the line connecting Earth to the center of the epicycle and the line connecting the center of uniform movement of the complex to the center of the epicycle (on the diagram that is angle DEB) cannot be easily derived.
There is one other benefit brought about by Ptolemy’s data selection. By choosing the data of mean sun being at quadrature, after the said angle is determined, he could directly apply the chord table to further determine the eccentricity in relation to the linear distance between Earth and the center of the epicycle. Since we also could calculate the ratio between the radius of the epicycle and the distance between Earth and the center of the epicycle, we could determine the eccentricity in relation to the radius of the epicycle. And previously we have determined the eccentricity between the center of the deferent and Earth in relation to the radius of the epicycle, we again could determine the relationship between the two eccentricities. We shall see that the later eccentricity is roughly double the earlier eccentricity. And it is in this way, by these procedures, and under such assumptions, that Ptolemy succeeds in determining those magnitudes related to the movement of Venus.
本文由 AI 从英文原文翻译。请以原文为准。
为了完成他所规定的数学家的任务,托勒密必须借助匀速圆周运动,解释金星所呈现的那些表观异常。在第九卷第二章,这两种异常被刻画为:“其中一种异常表现为与太阳有关,另一种则与黄道的各部分有关。”稍后分析金星运动时,这个初步区分得到了更清晰、更审慎的说明。眼睛所观察到的,首先是金星有时出现在太阳西侧,有时出现在东侧,似乎遵循某种周期模式;其次,金星与平太阳之间可能达到的最大角距离,也就是最大距角,会随着这个复合体(金星与太阳的相对关系)所出现的位置而变化;这里,视太阳与平太阳之间的差别可以忽略。
换句话说,第一种异常涉及这样一个事实:金星以与太阳相等的平均速度绕地球运行,却有时超过太阳,有时落在后面。第二种异常则是说,在这个复合体所运行的轨道上,若取两个不同位置,金星在那里与太阳之间所能达到的最大距角也会不同。这里所谓“可能”的最大距角,意思是:随着复合体不断绕地球运行,它每完成一个周期,便会到达同一个位置;而每次复合体到达这一位置时,金星在本轮上相对于太阳的位置又会改变,最终会到达本轮上的某个位置,使它在眼睛看来离太阳最远,这一位置便构成了可能达到的最大距角。但是,要使第二种异常成立,就必须先假定:在任何位置,都存在金星与太阳之间可能达到的最大距角。这是一个托勒密没有审慎确立的隐含假定。因为,如果让金星在本轮上的运行速度与本轮本身在均轮上的运行速度相同,我们就会看到,除了两个位置以外,再没有哪个位置能让金星达到它与太阳之间可能的最大距角;相反,当复合体到达均轮上的某一点时,金星在本轮上相对于太阳的位置也应是确定的。或者,如果复合体在均轮上的运行速度是金星在本轮上运行速度的某个倍数,也不可能在任意给定的位置都存在一个可能的最大距角。因而,后面的论述依赖于这样一个庞大的、托勒密没有澄清的假定。
在这两种异常得到澄清之后,似乎可以立即用托勒密解释太阳运动时所使用的现有假设来说明它们。然而,托勒密在第九卷第五章说:“一旦把观测到的具体位置应用于由这两个假设组合而构造的轨道,并不断将它们放在一起考察,我们就发现,事情无法如此简单地进行。”他从这些观测中发现的主要问题在于:即使将本轮假设与偏心圆假设结合起来,似乎仍有另一种异常有待解决。
这两个假设的结合如何解决前两种异常?金星有时在太阳西侧、有时在太阳东侧,可以由金星自身在本轮上的运动来解释。由于本轮中心以与平太阳相等的速度运动,而平太阳与视太阳相距不远,我们可以说,金星看起来在太阳西侧,是因为它运动到了本轮的一边;而当它出现在东侧,则是因为它运动到了本轮的另一边。第二种异常可以用偏心圆假设解释:按照这个假设,地球稍稍偏离复合体所运行之圆的圆心。其结果是,均轮的一侧比另一侧离地球更远。因此,合理地说,当复合体在均轮的一侧时,在观测者看来会比它处在另一侧时更大。相应地,在其中一侧,金星与太阳之间可能达到的最大距角,也会比另一侧可能达到的最大距角更大。通过这两个假设的结合,前两种异常得到了解决。
在我看来,第三种异常是在引入这个组合模型以后才产生的。托勒密把它说成一个从反复尝试中得到的发现,却没有明确界定它的确切性质。从解决办法,也就是等距点的引入,给这个组合模型带来的后果反推,我们可以说,第三种异常是复合体在均轮上的运动速率并不均匀。然而,若没有预先确立这个组合模型——复合体本身正是由它构造出来的——这个异常就无法得到充分表达。
现在,前两种异常,加上新产生的第三种异常,都可以在理论上得到解决。因此,要完成对金星运动的说明,剩下的工作便是在观测中找到出发点。首先需要的,就是一组记录不同位置上各种可能的最大距角的数据。
因为,从几何学的研究中我们知道,当金星位于从地球向本轮所作的切线与本轮的交点时,它与太阳之间便处于可能的最大距角,因为只有相切时,角距离才会最大。我们也知道,从几何学上说,复合体离地球越近,它在观测者看来可能达到的最大距角便越大。它在近地点离地球最近,在远地点则最远。因此,我们可以通过寻找所有可能的最大距角中最大的那个出现在哪里,来确定均轮上的远地点与近地点。拱线既已确定,所有数据便也能够安排到模型上。而且,借助已经从观测中得到的数据,也可以确定本轮半径与均轮半径之间的比例。因为,只要沿着托勒密的几何证明步骤,我们就能够确定地球至本轮中心的直线距离、均轮直径与本轮半径之间各自的关系。由此,也能够确定均轮中心与地球之间的偏心距。这个过程有趣的一点在于:托勒密协调分别从复合体处于远地点与近地点时所获得的结果时,采用了一种我们或许可以暂称为代数式的方法。他先假定某个东西——地球至本轮中心的直线距离——为已知,从而使所求之物——本轮半径——成为已知;当两边的结果不同时,他便通过建立两个拥有共同项的比例来协调它们。但我们在这里必须谨慎,要看到他的方法恰恰不是代数式的:他并不把所求之物假定为已知,而只是对出发点作出假定。换句话说,他的程序仍然是标准的:坚持先提出假设,再从假设出发达到结果这一形式,而不是把结果设定为已经知道的东西。
接下来,托勒密着手确定等距点——也就是金星绕之作匀速运动的中心——与地球之间的偏心距。他通过选择一个特定位置来做到这一点。正如他在第十卷第三章所说,这个位置是“太阳的平均位置在任一方向上距远地点四分之一圆周的时候”。然而,在收集了观测数据之后、开始证明之前,他说:“对于所提出的平均位置,最大的晨距角被假定为……,而最大的昏距角为……”(强调与省略为笔者所加。)
他借此处理的是最大晨距角与最大昏距角看起来不相等的问题。这种表观上的不相等,来自将由等距点解决的第三种异常。应当注意,他在这里对早先观测者所收集的数据的解释,似乎是:他们所假定的平太阳是以地球为运动中心的,而不是以均轮中心为运动中心。因此,他便方便地得出结论:当他模型中的平太阳运动到距近地点四分之一圆周时,前人所假定的平太阳也恰好相对于地球成直角。换句话说,以托勒密自己的模型为背景,早先天文学家所收集的数据便显得具有异常,而这源自他们隐含的假定:平太阳绕地球作匀速运动。只有通过这一解释,托勒密才能进行接下来的计算:把两个距角相加,构成一个更大的角,再把它分成两个相等的角,其中每一个都是距角。又或者,这并不是他的“解释”,而是他为了减轻自己的工作,特意选取了那些他知道持有这一假定的人所收集的数据。否则,如果已经假定平太阳在偏心圆上运行,那么,所假定的平太阳绕地球的运动,与符合他模型的平太阳绕自身运动中心的运动,二者之间便会失去直接的平行关系。(在塔利亚费罗版本第315页的图中,这个平行关系存在于线段DE与虚线之间。)没有这个直接的平行关系,就无法容易地求出连接地球与本轮中心的直线,同连接复合体匀速运动中心与本轮中心的直线之间的夹角,也就是图中的角DEB。
托勒密对数据的选择还带来了另一项好处。选取平太阳处于方照位置的数据以后,在上述角被确定后,他便能够直接运用弦表,进一步确定偏心距与地球至本轮中心的直线距离之间的关系。既然我们也能计算本轮半径与地球至本轮中心的距离之比,就可以确定偏心距与本轮半径之间的关系。而我们先前已经确定了均轮中心与地球之间的偏心距同本轮半径之间的关系,因此又可以确定这两个偏心距之间的关系。我们会看到,后一个偏心距大约是前一个的两倍。正是以这种方式,通过这些程序,并在这些假定之下,托勒密成功地确定了与金星运动相关的那些量。